diff --git a/.github/PULL_REQUEST_TEMPLATE.md b/.github/PULL_REQUEST_TEMPLATE.md
index 56cbd186..bb946ef2 100644
--- a/.github/PULL_REQUEST_TEMPLATE.md
+++ b/.github/PULL_REQUEST_TEMPLATE.md
@@ -45,4 +45,4 @@ Please describe the tests that you ran to verify your changes. Provide instructi
- [ ] My changes generate no new warnings
- [ ] I have added tests that prove my fix is effective or that my feature works
- [ ] New and existing unit tests pass locally with my changes
-- [ ] Any dependent changes have been merged and published in downstream modules
\ No newline at end of file
+- [ ] Any dependent changes have been merged and published in downstream modules
diff --git a/.gitignore b/.gitignore
index 26133bb2..fa94582a 100644
--- a/.gitignore
+++ b/.gitignore
@@ -104,7 +104,7 @@ ENV/
# Rope project settings
.ropeproject
-# DS Store
+# DS Store
.DS_Store
# for installation purposes
@@ -114,4 +114,4 @@ install-dev
/catboost_info/
# Data
-shapash/data/telco_customer_churn.csv
\ No newline at end of file
+shapash/data/telco_customer_churn.csv
diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
new file mode 100644
index 00000000..29ae367f
--- /dev/null
+++ b/.pre-commit-config.yaml
@@ -0,0 +1,70 @@
+default_language_version:
+ python: python3
+repos:
+- repo: https://github.com/pre-commit/pre-commit-hooks
+ rev: v3.2.0
+ hooks:
+ - id: check-ast
+ - id: check-byte-order-marker
+ - id: check-case-conflict
+ - id: check-docstring-first
+ - id: check-executables-have-shebangs
+ - id: check-json
+ - id: check-yaml
+ exclude: ^chart/
+ - id: debug-statements
+ - id: end-of-file-fixer
+ exclude: ^(docs/|gdocs/)
+ - id: pretty-format-json
+ args: ['--autofix']
+ exclude: .ipynb
+ - id: trailing-whitespace
+ args: ['--markdown-linebreak-ext=md']
+ exclude: ^(docs/|gdocs/)
+ - id: mixed-line-ending
+ args: ['--fix=lf']
+ exclude: ^(docs/|gdocs/)
+ - id: check-added-large-files
+ args: ['--maxkb=500']
+ - id: no-commit-to-branch
+ args: ['--branch', 'master', '--branch', 'develop']
+- repo: https://github.com/psf/black
+ rev: 21.12b0
+ hooks:
+ - id: black
+ args: [--line-length=120]
+ additional_dependencies: ['click==8.0.4']
+#- repo: https://github.com/pre-commit/mirrors-mypy
+# rev: 'v0.931'
+# hooks:
+# - id: mypy
+# args: [--ignore-missing-imports, --disallow-untyped-defs, --show-error-codes, --no-site-packages]
+# files: src
+# - repo: https://github.com/PyCQA/flake8
+# rev: 6.0.0
+# hooks:
+# - id: flake8
+# exclude: ^tests/
+# args: ['--ignore=E501,D2,D3,D4,D104,D100,D106,D107,W503,D105,E203']
+# additional_dependencies: [ flake8-docstrings, "flake8-bugbear==22.8.23" ]
+- repo: https://github.com/pre-commit/mirrors-isort
+ rev: v5.4.2
+ hooks:
+ - id: isort
+ args: ["--profile", "black", "-l", "120"]
+- repo: https://github.com/asottile/pyupgrade
+ rev: v2.7.2
+ hooks:
+ - id: pyupgrade
+ args: [--py37-plus]
+- repo: https://github.com/asottile/blacken-docs
+ rev: v1.8.0
+ hooks:
+ - id: blacken-docs
+ additional_dependencies: [black==21.12b0]
+- repo: https://github.com/compilerla/conventional-pre-commit
+ rev: v2.1.1
+ hooks:
+ - id: conventional-pre-commit
+ stages: [commit-msg]
+ args: [] # optional: list of Conventional Commits types to allow e.g. [feat, fix, ci, chore, test]
diff --git a/.readthedocs.yml b/.readthedocs.yml
index f44b00d3..0c09bd5e 100644
--- a/.readthedocs.yml
+++ b/.readthedocs.yml
@@ -21,7 +21,7 @@ build:
os: ubuntu-20.04
tools:
python: "3.10"
-
+
# Optionally set the version of Python and requirements required to build your docs
python:
install:
diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md
index 0e7bc96f..7a650ec8 100644
--- a/CONTRIBUTING.md
+++ b/CONTRIBUTING.md
@@ -58,7 +58,7 @@ To contribute to Shapash, you will need to create a personal branch.
```
git checkout -b feature/my-contribution-branch
```
-We recommand to use a convention of naming branch.
+We recommand to use a convention of naming branch.
- **feature/your_feature_name** if you are creating a feature
- **hotfix/your_bug_fix** if you are fixing a bug
@@ -70,7 +70,7 @@ Before committing your modifications, we have some recommendations :
```
pytest
```
-- Try to build Shapash
+- Try to build Shapash
```
python setup.py bdist_wheel
```
@@ -91,7 +91,7 @@ git commit -m ‘fixed a bug’
git push origin feature/my-contribution-branch
```
-Your branch is now available on your remote forked repository, with your changes.
+Your branch is now available on your remote forked repository, with your changes.
Next step is now to create a Pull Request so the Shapash Team can add your changes to the official repository.
@@ -104,7 +104,7 @@ To create one, on the top of your forked repository, you will find a button "Com
@@ -145,13 +145,13 @@ Shapash can use category-encoders object, sklearn ColumnTransformer or simply fe
Shapash is intended to work with Python versions 3.8 to 3.11. Installation can be done with pip:
-```python
+```bash
pip install shapash
```
In order to generate the Shapash Report some extra requirements are needed.
You can install these using the following command :
-```python
+```bash
pip install shapash[report]
```
@@ -167,24 +167,25 @@ The 4 steps to display results:
```python
from shapash import SmartExplainer
+
xpl = SmartExplainer(
- model=regressor,
- features_dict=house_dict, # Optional parameter
- preprocessing=encoder, # Optional: compile step can use inverse_transform method
- postprocessing=postprocess, # Optional: see tutorial postprocessing
+ model=regressor,
+ features_dict=house_dict, # Optional parameter
+ preprocessing=encoder, # Optional: compile step can use inverse_transform method
+ postprocessing=postprocess, # Optional: see tutorial postprocessing
)
```
- Step 2: Compile Dataset, ...
> There 1 mandatory parameter in compile method: Dataset
-
+
```python
xpl.compile(
- x=xtest,
- y_pred=y_pred, # Optional: for your own prediction (by default: model.predict)
- y_target=yTest, # Optional: allows to display True Values vs Predicted Values
- additional_data=xadditional, # Optional: additional dataset of features for Webapp
- additional_features_dict=features_dict_additional, # Optional: dict additional data
+ x=xtest,
+ y_pred=y_pred, # Optional: for your own prediction (by default: model.predict)
+ y_target=yTest, # Optional: allows to display True Values vs Predicted Values
+ additional_data=xadditional, # Optional: additional dataset of features for Webapp
+ additional_features_dict=features_dict_additional, # Optional: dict additional data
)
```
@@ -193,7 +194,7 @@ xpl.compile(
```python
app = xpl.run_app()
-```
+```
[Live Demo Shapash-Monitor](https://shapash-demo.ossbymaif.fr/)
@@ -203,15 +204,15 @@ app = xpl.run_app()
```python
xpl.generate_report(
- output_file='path/to/output/report.html',
- project_info_file='path/to/project_info.yml',
+ output_file="path/to/output/report.html",
+ project_info_file="path/to/project_info.yml",
x_train=xtrain,
y_train=ytrain,
y_test=ytest,
title_story="House prices report",
title_description="""This document is a data science report of the kaggle house prices tutorial project.
It was generated using the Shapash library.""",
- metrics=[{'name': 'MSE', 'path': 'sklearn.metrics.mean_squared_error'}]
+ metrics=[{"name": "MSE", "path": "sklearn.metrics.mean_squared_error"}],
)
```
@@ -220,9 +221,9 @@ xpl.generate_report(
- Step 5: From training to deployment : SmartPredictor Object
> Shapash provides a SmartPredictor object to deploy the summary of local explanation for the operational needs.
It is an object dedicated to deployment, lighter than SmartExplainer with additional consistency checks.
- SmartPredictor can be used with an API or in batch mode. It provides predictions, detailed or summarized local
+ SmartPredictor can be used with an API or in batch mode. It provides predictions, detailed or summarized local
explainability using appropriate wording.
-
+
```python
predictor = xpl.to_smartpredictor()
```
diff --git a/data/house_prices_labels.json b/data/house_prices_labels.json
index cda51b24..1f3e13da 100644
--- a/data/house_prices_labels.json
+++ b/data/house_prices_labels.json
@@ -1,74 +1,74 @@
{
- "MSSubClass": "Building Class",
- "MSZoning": "General zoning classification",
- "LotArea": "Lot size square feet",
- "Street": "Type of road access",
- "LotShape": "General shape of property",
- "LandContour": "Flatness of the property",
- "Utilities": "Type of utilities available",
- "LotConfig": "Lot configuration",
- "LandSlope": "Slope of property",
- "Neighborhood": "Physical locations within Ames city limits",
- "Condition1": "Proximity to various conditions",
- "Condition2": "Proximity to other various conditions",
- "BldgType": "Type of dwelling",
- "HouseStyle": "Style of dwelling",
- "OverallQual": "Overall material and finish of the house",
- "OverallCond": "Overall condition of the house",
- "YearBuilt": "Original construction date",
- "YearRemodAdd": "Remodel date",
- "RoofStyle": "Type of roof",
- "RoofMatl": "Roof material",
- "Exterior1st": "Exterior covering on house",
- "Exterior2nd": "Other exterior covering on house",
- "MasVnrType": "Masonry veneer type",
- "MasVnrArea": "Masonry veneer area in square feet",
- "ExterQual": "Exterior materials' quality",
- "ExterCond": "Exterior materials' condition",
- "Foundation": "Type of foundation",
- "BsmtQual": "Height of the basement",
- "BsmtCond": "General condition of the basement",
- "BsmtExposure": "Refers to walkout or garden level walls",
- "BsmtFinType1": "Rating of basement finished area",
- "BsmtFinSF1": "Type 1 finished square feet",
- "BsmtFinType2": "Rating of basement finished area (if present)",
- "BsmtFinSF2": "Type 2 finished square feet",
- "BsmtUnfSF": "Unfinished square feet of basement area",
- "TotalBsmtSF": "Total square feet of basement area",
- "Heating": "Type of heating",
- "HeatingQC": "Heating quality and condition",
- "CentralAir": "Central air conditioning",
- "Electrical": "Electrical system",
- "1stFlrSF": "First Floor square feet",
- "2ndFlrSF": "Second floor square feet",
- "LowQualFinSF": "Low quality finished square feet",
- "GrLivArea": "Ground living area square feet",
- "BsmtFullBath": "Basement full bathrooms",
- "BsmtHalfBath": "Basement half bathrooms",
- "FullBath": "Full bathrooms above grade",
- "HalfBath": "Half baths above grade",
- "BedroomAbvGr": "Bedrooms above grade",
- "KitchenAbvGr": "Kitchens above grade",
- "KitchenQual": "Kitchen quality",
- "TotRmsAbvGrd": "Total rooms above grade",
- "Functional": "Home functionality",
- "Fireplaces": "Number of fireplaces",
- "GarageType": "Garage location",
- "GarageYrBlt": "Year garage was built",
- "GarageFinish": "Interior finish of the garage?",
- "GarageArea": "Size of garage in square feet",
- "GarageQual": "Garage quality",
- "GarageCond": "Garage condition",
- "PavedDrive": "Paved driveway",
- "WoodDeckSF": "Wood deck area in square feet",
- "OpenPorchSF": "Open porch area in square feet",
- "EnclosedPorch": "Enclosed porch area in square feet",
- "3SsnPorch": "Three season porch area in square feet",
- "ScreenPorch": "Screen porch area in square feet",
- "PoolArea": "Pool area in square feet",
- "MiscVal": "$Value of miscellaneous feature",
- "MoSold": "Month Sold",
- "YrSold": "Year Sold",
- "SaleType": "Type of sale",
- "SaleCondition": "Condition of sale"
-}
\ No newline at end of file
+ "1stFlrSF": "First Floor square feet",
+ "2ndFlrSF": "Second floor square feet",
+ "3SsnPorch": "Three season porch area in square feet",
+ "BedroomAbvGr": "Bedrooms above grade",
+ "BldgType": "Type of dwelling",
+ "BsmtCond": "General condition of the basement",
+ "BsmtExposure": "Refers to walkout or garden level walls",
+ "BsmtFinSF1": "Type 1 finished square feet",
+ "BsmtFinSF2": "Type 2 finished square feet",
+ "BsmtFinType1": "Rating of basement finished area",
+ "BsmtFinType2": "Rating of basement finished area (if present)",
+ "BsmtFullBath": "Basement full bathrooms",
+ "BsmtHalfBath": "Basement half bathrooms",
+ "BsmtQual": "Height of the basement",
+ "BsmtUnfSF": "Unfinished square feet of basement area",
+ "CentralAir": "Central air conditioning",
+ "Condition1": "Proximity to various conditions",
+ "Condition2": "Proximity to other various conditions",
+ "Electrical": "Electrical system",
+ "EnclosedPorch": "Enclosed porch area in square feet",
+ "ExterCond": "Exterior materials' condition",
+ "ExterQual": "Exterior materials' quality",
+ "Exterior1st": "Exterior covering on house",
+ "Exterior2nd": "Other exterior covering on house",
+ "Fireplaces": "Number of fireplaces",
+ "Foundation": "Type of foundation",
+ "FullBath": "Full bathrooms above grade",
+ "Functional": "Home functionality",
+ "GarageArea": "Size of garage in square feet",
+ "GarageCond": "Garage condition",
+ "GarageFinish": "Interior finish of the garage?",
+ "GarageQual": "Garage quality",
+ "GarageType": "Garage location",
+ "GarageYrBlt": "Year garage was built",
+ "GrLivArea": "Ground living area square feet",
+ "HalfBath": "Half baths above grade",
+ "Heating": "Type of heating",
+ "HeatingQC": "Heating quality and condition",
+ "HouseStyle": "Style of dwelling",
+ "KitchenAbvGr": "Kitchens above grade",
+ "KitchenQual": "Kitchen quality",
+ "LandContour": "Flatness of the property",
+ "LandSlope": "Slope of property",
+ "LotArea": "Lot size square feet",
+ "LotConfig": "Lot configuration",
+ "LotShape": "General shape of property",
+ "LowQualFinSF": "Low quality finished square feet",
+ "MSSubClass": "Building Class",
+ "MSZoning": "General zoning classification",
+ "MasVnrArea": "Masonry veneer area in square feet",
+ "MasVnrType": "Masonry veneer type",
+ "MiscVal": "$Value of miscellaneous feature",
+ "MoSold": "Month Sold",
+ "Neighborhood": "Physical locations within Ames city limits",
+ "OpenPorchSF": "Open porch area in square feet",
+ "OverallCond": "Overall condition of the house",
+ "OverallQual": "Overall material and finish of the house",
+ "PavedDrive": "Paved driveway",
+ "PoolArea": "Pool area in square feet",
+ "RoofMatl": "Roof material",
+ "RoofStyle": "Type of roof",
+ "SaleCondition": "Condition of sale",
+ "SaleType": "Type of sale",
+ "ScreenPorch": "Screen porch area in square feet",
+ "Street": "Type of road access",
+ "TotRmsAbvGrd": "Total rooms above grade",
+ "TotalBsmtSF": "Total square feet of basement area",
+ "Utilities": "Type of utilities available",
+ "WoodDeckSF": "Wood deck area in square feet",
+ "YearBuilt": "Original construction date",
+ "YearRemodAdd": "Remodel date",
+ "YrSold": "Year Sold"
+}
diff --git a/data/titaniclabels.json b/data/titaniclabels.json
index 7a0c8e1e..7aff00fd 100644
--- a/data/titaniclabels.json
+++ b/data/titaniclabels.json
@@ -1 +1,13 @@
-{"PassengerID": "PassengerID", "Survival": "Has survived ?", "Pclass": "Ticket class", "Name": "Name, First name", "Sex": "Sex", "Age": "Age", "SibSp": "Relatives such as brother or wife", "Parch": "Relatives like children or parents", "Fare": "Passenger fare", "Embarked": "Port of embarkation", "Title": "Title of passenger"}
\ No newline at end of file
+{
+ "Age": "Age",
+ "Embarked": "Port of embarkation",
+ "Fare": "Passenger fare",
+ "Name": "Name, First name",
+ "Parch": "Relatives like children or parents",
+ "PassengerID": "PassengerID",
+ "Pclass": "Ticket class",
+ "Sex": "Sex",
+ "SibSp": "Relatives such as brother or wife",
+ "Survival": "Has survived ?",
+ "Title": "Title of passenger"
+}
diff --git a/docs/conf.py b/docs/conf.py
index b0f2c98c..6dfb8a60 100644
--- a/docs/conf.py
+++ b/docs/conf.py
@@ -12,15 +12,16 @@
#
import os
import sys
-sys.path.insert(0, '..')
+
+sys.path.insert(0, "..")
import shapash
# -- Project information -----------------------------------------------------
-project = 'Shapash'
-copyright = '2020, Maif'
-author = 'Maif'
+project = "Shapash"
+copyright = "2020, Maif"
+author = "Maif"
# The short X.Y version
version = shapash.__version__
@@ -35,30 +36,30 @@
# extensions coming with Sphinx (named 'sphinx.ext.*') or your custom
# ones.
extensions = [
- 'sphinx.ext.autodoc',
- 'sphinx.ext.viewcode',
- 'sphinx.ext.todo',
- 'sphinx.ext.napoleon',
- 'nbsphinx',
+ "sphinx.ext.autodoc",
+ "sphinx.ext.viewcode",
+ "sphinx.ext.todo",
+ "sphinx.ext.napoleon",
+ "nbsphinx",
]
-nbsphinx_execute = 'never'
-master_doc = 'index'
+nbsphinx_execute = "never"
+master_doc = "index"
# Add any paths that contain templates here, relative to this directory.
-templates_path = ['_templates']
+templates_path = ["_templates"]
# The language for content autogenerated by Sphinx. Refer to documentation
# for a list of supported languages.
#
# This is also used if you do content translation via gettext catalogs.
# Usually you set "language" from the command line for these cases.
-language = 'en'
+language = "en"
# List of patterns, relative to source directory, that match files and
# directories to ignore when looking for source files.
# This pattern also affects html_static_path and html_extra_path.
-exclude_patterns = ['_build', 'Thumbs.db', '.DS_Store']
+exclude_patterns = ["_build", "Thumbs.db", ".DS_Store"]
# -- Options for HTML output -------------------------------------------------
@@ -66,41 +67,35 @@
# The theme to use for HTML and HTML Help pages. See the documentation for
# a list of builtin themes.
#
-html_theme = 'sphinx_material'
-#html_logo = './assets/images/svg/shapash-github.svg'
+html_theme = "sphinx_material"
+# html_logo = './assets/images/svg/shapash-github.svg'
# Material theme options (see theme.conf for more information)
html_theme_options = {
-
# Set the name of the project to appear in the navigation.
- 'nav_title': 'Shapash',
+ "nav_title": "Shapash",
# Set the color and the accent color
- 'color_primary': 'amber',
- 'color_accent': 'deep-orange',
-
+ "color_primary": "amber",
+ "color_accent": "deep-orange",
# Set the repo location to get a badge with stats
- 'repo_url': 'https://github.com/MAIF/shapash',
- 'repo_name': 'shapash',
-
+ "repo_url": "https://github.com/MAIF/shapash",
+ "repo_name": "shapash",
# Icon of the navbar
- 'logo_icon': '',
-
+ "logo_icon": "",
# Visible levels of the global TOC; -1 means unlimited
- 'globaltoc_depth': 3,
+ "globaltoc_depth": 3,
# If False, expand all TOC entries
- 'globaltoc_collapse': True,
+ "globaltoc_collapse": True,
# If True, show hidden TOC entries
- 'globaltoc_includehidden': False,
+ "globaltoc_includehidden": False,
}
-html_sidebars = {
- "**": ["logo-text.html", "globaltoc.html", "localtoc.html", "searchbox.html"]
-}
+html_sidebars = {"**": ["logo-text.html", "globaltoc.html", "localtoc.html", "searchbox.html"]}
# Add any paths that contain custom static files (such as style sheets) here,
# relative to this directory. They are copied after the builtin static files,
# so a file named "default.css" will overwrite the builtin "default.css".
-html_static_path = ['_static']
+html_static_path = ["_static"]
# -- Extension configuration -------------------------------------------------
@@ -112,9 +107,10 @@
# -- Additional html pages -------------------------------------------------
import subprocess
+
# Generates the report example in the documentation
-subprocess.call(['python', '../tutorial/generate_report/shapash_report_example.py'])
-html_extra_path = ['../tutorial/report/output/report.html']
+subprocess.call(["python", "../tutorial/generate_report/shapash_report_example.py"])
+html_extra_path = ["../tutorial/report/output/report.html"]
def setup_tutorials():
@@ -123,33 +119,45 @@ def setup_tutorials():
def _copy_notebooks_and_create_rst(d_path, new_d_path, d_name):
os.makedirs(new_d_path, exist_ok=True)
- list_notebooks = [f for f in os.listdir(d_path) if os.path.splitext(f)[-1] == '.ipynb']
+ list_notebooks = [f for f in os.listdir(d_path) if os.path.splitext(f)[-1] == ".ipynb"]
for notebook_f_name in list_notebooks:
shutil.copyfile(os.path.join(d_path, notebook_f_name), os.path.join(new_d_path, notebook_f_name))
# RST file (see for example docs/overview.rst)
- rst_file = '\n'.join(
- [f'{d_name}', '======================', '', '.. toctree::',
- ' :maxdepth: 1', ' :glob:', '', ' *', '']
+ rst_file = "\n".join(
+ [
+ f"{d_name}",
+ "======================",
+ "",
+ ".. toctree::",
+ " :maxdepth: 1",
+ " :glob:",
+ "",
+ " *",
+ "",
+ ]
)
- with open(os.path.join(new_d_path, 'index.rst'), 'w') as f:
+ with open(os.path.join(new_d_path, "index.rst"), "w") as f:
f.write(rst_file)
docs_path = pathlib.Path(__file__).parent
- tutorials_path = os.path.join(docs_path.parent, 'tutorial')
- tutorials_doc_path = os.path.join(docs_path, 'tutorials')
+ tutorials_path = os.path.join(docs_path.parent, "tutorial")
+ tutorials_doc_path = os.path.join(docs_path, "tutorials")
# Create a directory in shapash/docs/tutorials for each directory of shapash/tutorial
# And copy each notebook file in it
- list_dir = [d for d in os.listdir(tutorials_path) if os.path.isdir(os.path.join(tutorials_path, d))
- and not d.startswith('.')]
+ list_dir = [
+ d
+ for d in os.listdir(tutorials_path)
+ if os.path.isdir(os.path.join(tutorials_path, d)) and not d.startswith(".")
+ ]
for d_name in list_dir:
d_path = os.path.join(tutorials_path, d_name)
new_d_path = os.path.join(tutorials_doc_path, d_name)
_copy_notebooks_and_create_rst(d_path, new_d_path, d_name)
# Also copying all the overview tutorials (shapash/tutorial/shapash-overview-in-jupyter.ipynb for example)
- _copy_notebooks_and_create_rst(tutorials_path, tutorials_doc_path, 'overview')
+ _copy_notebooks_and_create_rst(tutorials_path, tutorials_doc_path, "overview")
setup_tutorials()
diff --git a/requirements.dev.txt b/requirements.dev.txt
index 9dfe603e..6e89a7b2 100644
--- a/requirements.dev.txt
+++ b/requirements.dev.txt
@@ -25,7 +25,7 @@ nbsphinx==0.8.8
sphinx_material==0.0.35
pytest>=6.2.5
pytest-cov>=2.8.1
-scikit-learn>=1.0.1
+scikit-learn>=1.0.1,<1.4
xgboost>=1.0.0
nbformat>4.2.0
numba>=0.53.1
diff --git a/setup.py b/setup.py
index 11b1c462..76df371c 100644
--- a/setup.py
+++ b/setup.py
@@ -1,108 +1,121 @@
#!/usr/bin/env python
-# -*- coding: utf-8 -*-
"""The setup script."""
import os
+
from setuptools import setup
here = os.path.abspath(os.path.dirname(__file__))
-with open('README.md', encoding='utf8') as readme_file:
+with open("README.md", encoding="utf8") as readme_file:
long_description = readme_file.read()
# Load the package's __version__.py module as a dictionary.
version_d: dict = {}
-with open(os.path.join(here, 'shapash', "__version__.py")) as f:
+with open(os.path.join(here, "shapash", "__version__.py")) as f:
exec(f.read(), version_d)
requirements = [
- 'plotly>=5.0.0',
- 'matplotlib>=3.2.0',
- 'numpy>1.18.0',
- 'pandas>1.0.2',
- 'shap>=0.38.1',
- 'Flask<2.3.0',
- 'dash>=2.3.1',
- 'dash-bootstrap-components>=1.1.0',
- 'dash-core-components>=2.0.0',
- 'dash-daq>=0.5.0',
- 'dash-html-components>=2.0.0',
- 'dash-renderer==1.8.3',
- 'dash-table>=5.0.0',
- 'nbformat>4.2.0',
- 'numba>=0.53.1',
- 'scikit-learn>=1.0.1',
- 'category_encoders>=2.6.0',
- 'scipy>=0.19.1',
+ "plotly>=5.0.0",
+ "matplotlib>=3.2.0",
+ "numpy>1.18.0",
+ "pandas>1.0.2",
+ "shap>=0.38.1",
+ "Flask<2.3.0",
+ "dash>=2.3.1",
+ "dash-bootstrap-components>=1.1.0",
+ "dash-core-components>=2.0.0",
+ "dash-daq>=0.5.0",
+ "dash-html-components>=2.0.0",
+ "dash-renderer==1.8.3",
+ "dash-table>=5.0.0",
+ "nbformat>4.2.0",
+ "numba>=0.53.1",
+ "scikit-learn>=1.0.1,<1.4",
+ "category_encoders>=2.6.0",
+ "scipy>=0.19.1",
]
extras = dict()
# This list should be identical to the list in shapash/report/__init__.py
-extras['report'] = [
- 'nbconvert>=6.0.7',
- 'papermill>=2.0.0',
- 'jupyter-client>=7.4.0',
- 'seaborn==0.12.2',
- 'notebook',
- 'Jinja2>=2.11.0',
- 'phik'
+extras["report"] = [
+ "nbconvert>=6.0.7",
+ "papermill>=2.0.0",
+ "jupyter-client>=7.4.0",
+ "seaborn==0.12.2",
+ "notebook",
+ "Jinja2>=2.11.0",
+ "phik",
]
-extras['xgboost'] = ['xgboost>=1.0.0']
-extras['lightgbm'] = ['lightgbm>=2.3.0']
-extras['catboost'] = ['catboost>=1.0.1']
-extras['lime'] = ['lime>=0.2.0.0']
+extras["xgboost"] = ["xgboost>=1.0.0"]
+extras["lightgbm"] = ["lightgbm>=2.3.0"]
+extras["catboost"] = ["catboost>=1.0.1"]
+extras["lime"] = ["lime>=0.2.0.0"]
-setup_requirements = ['pytest-runner', ]
+setup_requirements = [
+ "pytest-runner",
+]
-test_requirements = ['pytest', ]
+test_requirements = [
+ "pytest",
+]
setup(
name="shapash",
- version=version_d['__version__'],
- python_requires='>3.7, <3.12',
- url='https://github.com/MAIF/shapash',
+ version=version_d["__version__"],
+ python_requires=">3.7, <3.12",
+ url="https://github.com/MAIF/shapash",
author="Yann Golhen, Sebastien Bidault, Yann Lagre, Maxime Gendre",
author_email="yann.golhen@maif.fr",
description="Shapash is a Python library which aims to make machine learning interpretable and understandable by everyone.",
long_description=long_description,
- long_description_content_type='text/markdown',
+ long_description_content_type="text/markdown",
classifiers=[
- 'Programming Language :: Python :: 3',
- 'Programming Language :: Python :: 3.8',
- 'Programming Language :: Python :: 3.9',
- 'Programming Language :: Python :: 3.10',
- 'Programming Language :: Python :: 3.11',
+ "Programming Language :: Python :: 3",
+ "Programming Language :: Python :: 3.8",
+ "Programming Language :: Python :: 3.9",
+ "Programming Language :: Python :: 3.10",
+ "Programming Language :: Python :: 3.11",
"License :: OSI Approved :: Apache Software License",
"Operating System :: OS Independent",
],
install_requires=requirements,
extras_require=extras,
license="Apache Software License 2.0",
- keywords='shapash',
+ keywords="shapash",
package_dir={
- 'shapash': 'shapash',
- 'shapash.data': 'shapash/data',
- 'shapash.decomposition': 'shapash/decomposition',
- 'shapash.explainer': 'shapash/explainer',
- 'shapash.backend': 'shapash/backend',
- 'shapash.manipulation': 'shapash/manipulation',
- 'shapash.report': 'shapash/report',
- 'shapash.utils': 'shapash/utils',
- 'shapash.webapp': 'shapash/webapp',
- 'shapash.webapp.utils': 'shapash/webapp/utils',
- 'shapash.style': 'shapash/style',
+ "shapash": "shapash",
+ "shapash.data": "shapash/data",
+ "shapash.decomposition": "shapash/decomposition",
+ "shapash.explainer": "shapash/explainer",
+ "shapash.backend": "shapash/backend",
+ "shapash.manipulation": "shapash/manipulation",
+ "shapash.report": "shapash/report",
+ "shapash.utils": "shapash/utils",
+ "shapash.webapp": "shapash/webapp",
+ "shapash.webapp.utils": "shapash/webapp/utils",
+ "shapash.style": "shapash/style",
},
- packages=['shapash', 'shapash.data', 'shapash.decomposition',
- 'shapash.explainer', 'shapash.backend', 'shapash.manipulation',
- 'shapash.utils', 'shapash.webapp', 'shapash.webapp.utils',
- 'shapash.report', 'shapash.style'],
- data_files=[('style', ['shapash/style/colors.json'])],
+ packages=[
+ "shapash",
+ "shapash.data",
+ "shapash.decomposition",
+ "shapash.explainer",
+ "shapash.backend",
+ "shapash.manipulation",
+ "shapash.utils",
+ "shapash.webapp",
+ "shapash.webapp.utils",
+ "shapash.report",
+ "shapash.style",
+ ],
+ data_files=[("style", ["shapash/style/colors.json"])],
include_package_data=True,
setup_requires=setup_requirements,
- test_suite='tests',
+ test_suite="tests",
tests_require=test_requirements,
zip_safe=False,
)
diff --git a/shapash/__init__.py b/shapash/__init__.py
index 52cbbfb2..59117772 100644
--- a/shapash/__init__.py
+++ b/shapash/__init__.py
@@ -1,8 +1,10 @@
-
"""Top-level package."""
-__author__ = """Yann Golhen, Yann Lagré, Sebastien Bidault, Maxime Gendre, Thomas Bouche, Johann Martin, Guillaume Vignal"""
-__email__ = 'yann.golhen@maif.fr, yann.lagre@maif.fr, sebabstien.bidault.marketing@maif.fr, thomas.bouche@maif.fr, guillaume.vignal@maif.fr'
+__author__ = (
+ """Yann Golhen, Yann Lagré, Sebastien Bidault, Maxime Gendre, Thomas Bouche, Johann Martin, Guillaume Vignal"""
+)
+__email__ = "yann.golhen@maif.fr, yann.lagre@maif.fr, sebabstien.bidault.marketing@maif.fr, thomas.bouche@maif.fr, guillaume.vignal@maif.fr"
-from .__version__ import __version__
from shapash.explainer.smart_explainer import SmartExplainer
+
+from .__version__ import __version__
diff --git a/shapash/backend/__init__.py b/shapash/backend/__init__.py
index b9d262ab..0627fb74 100644
--- a/shapash/backend/__init__.py
+++ b/shapash/backend/__init__.py
@@ -1,9 +1,9 @@
-import sys
import inspect
+import sys
from .base_backend import BaseBackend
-from .shap_backend import ShapBackend
from .lime_backend import LimeBackend
+from .shap_backend import ShapBackend
def get_backend_cls_from_name(name):
@@ -14,10 +14,10 @@ def get_backend_cls_from_name(name):
cls
for _, cls in inspect.getmembers(sys.modules[__name__])
if (
- inspect.isclass(cls)
- and issubclass(cls, BaseBackend)
- and cls.name.lower() == name.lower()
- and cls.name.lower() != 'base'
+ inspect.isclass(cls)
+ and issubclass(cls, BaseBackend)
+ and cls.name.lower() == name.lower()
+ and cls.name.lower() != "base"
)
]
diff --git a/shapash/backend/base_backend.py b/shapash/backend/base_backend.py
index cd7ef5b8..6e3699a3 100644
--- a/shapash/backend/base_backend.py
+++ b/shapash/backend/base_backend.py
@@ -1,9 +1,10 @@
from abc import ABC, abstractmethod
from typing import Any, List, Optional, Union
-import pandas as pd
+
import numpy as np
+import pandas as pd
-from shapash.utils.check import check_model, check_contribution_object
+from shapash.utils.check import check_contribution_object, check_model
from shapash.utils.transform import adapt_contributions, get_preprocessing_mapping
from shapash.utils.utils import choose_state
@@ -19,13 +20,13 @@ class BaseBackend(ABC):
# `column_aggregation` defines a way to aggregate local contributions.
# Default is sum, possible values are 'sum' or 'first'.
# It allows to compute (column-wise) aggregation of local contributions.
- column_aggregation = 'sum'
+ column_aggregation = "sum"
# `name` defines the string name of the backend allowing to identify and
# construct the backend from it.
- name = 'base'
+ name = "base"
support_groups = True
- supported_cases = ['classification', 'regression']
+ supported_cases = ["classification", "regression"]
def __init__(self, model: Any, preprocessing: Optional[Any] = None):
"""Create a backend instance using a given implementation.
@@ -43,7 +44,7 @@ def __init__(self, model: Any, preprocessing: Optional[Any] = None):
self.state = None
self._case, self._classes = check_model(model)
if self._case not in self.supported_cases:
- raise ValueError(f'Model not supported by the backend as it does not cover {self._case} case')
+ raise ValueError(f"Model not supported by the backend as it does not cover {self._case} case")
@abstractmethod
def run_explainer(self, x: pd.DataFrame) -> dict:
@@ -53,10 +54,7 @@ def run_explainer(self, x: pd.DataFrame) -> dict:
)
def get_local_contributions(
- self,
- x: pd.DataFrame,
- explain_data: Any,
- subset: Optional[List[int]] = None
+ self, x: pd.DataFrame, explain_data: Any, subset: Optional[List[int]] = None
) -> Union[pd.DataFrame, List[pd.DataFrame]]:
"""Get local contributions using the explainer data computed in the `run_explainer`
method.
@@ -81,24 +79,21 @@ def get_local_contributions(
The local contributions computed by the backend.
"""
assert isinstance(explain_data, dict), "The _run_explainer method should return a dict"
- if 'contributions' not in explain_data.keys():
+ if "contributions" not in explain_data.keys():
raise ValueError(
- 'The _run_explainer method should return a dict'
- ' with at least `contributions` key containing '
- 'the local contributions'
+ "The _run_explainer method should return a dict"
+ " with at least `contributions` key containing "
+ "the local contributions"
)
- local_contributions = explain_data['contributions']
+ local_contributions = explain_data["contributions"]
if subset is not None:
local_contributions = local_contributions.loc[subset]
local_contributions = self.format_and_aggregate_local_contributions(x, local_contributions)
return local_contributions
def get_global_features_importance(
- self,
- contributions: pd.DataFrame,
- explain_data: Optional[dict] = None,
- subset: Optional[List[int]] = None
+ self, contributions: pd.DataFrame, explain_data: Optional[dict] = None, subset: Optional[List[int]] = None
) -> Union[pd.Series, List[pd.Series]]:
"""Get global contributions using the explainer data computed in the `run_explainer`
method.
@@ -126,9 +121,9 @@ def get_global_features_importance(
return state.compute_features_import(contributions)
def format_and_aggregate_local_contributions(
- self,
- x: pd.DataFrame,
- contributions: Union[pd.DataFrame, np.array, List[pd.DataFrame], List[np.array]],
+ self,
+ x: pd.DataFrame,
+ contributions: Union[pd.DataFrame, np.array, List[pd.DataFrame], List[np.array]],
) -> Union[pd.DataFrame, List[pd.DataFrame]]:
"""
This function allows to format and aggregate contributions in the right format
@@ -153,7 +148,8 @@ def format_and_aggregate_local_contributions(
check_contribution_object(self._case, self._classes, contributions)
contributions = self.state.validate_contributions(contributions, x)
contributions_cols = (
- contributions.columns.to_list() if isinstance(contributions, pd.DataFrame)
+ contributions.columns.to_list()
+ if isinstance(contributions, pd.DataFrame)
else contributions[0].columns.to_list()
)
if _needs_preprocessing(contributions_cols, x, self.preprocessing):
@@ -161,8 +157,7 @@ def format_and_aggregate_local_contributions(
return contributions
def _apply_preprocessing(
- self,
- contributions: Union[pd.DataFrame, List[pd.DataFrame]]
+ self, contributions: Union[pd.DataFrame, List[pd.DataFrame]]
) -> Union[pd.DataFrame, List[pd.DataFrame]]:
"""
Reconstruct contributions for original features, taken into account a preprocessing.
@@ -179,9 +174,7 @@ def _apply_preprocessing(
"""
if self.preprocessing:
return self.state.inverse_transform_contributions(
- contributions,
- self.preprocessing,
- agg_columns=self.column_aggregation
+ contributions, self.preprocessing, agg_columns=self.column_aggregation
)
else:
return contributions
diff --git a/shapash/backend/lime_backend.py b/shapash/backend/lime_backend.py
index ada25113..3830ba21 100644
--- a/shapash/backend/lime_backend.py
+++ b/shapash/backend/lime_backend.py
@@ -1,10 +1,11 @@
try:
from lime import lime_tabular
+
is_lime_available = True
except ImportError:
is_lime_available = False
-from typing import Any, Optional, List, Union
+from typing import Any, List, Optional, Union
import pandas as pd
@@ -12,12 +13,12 @@
class LimeBackend(BaseBackend):
- column_aggregation = 'sum'
- name = 'lime'
+ column_aggregation = "sum"
+ name = "lime"
support_groups = False
def __init__(self, model, preprocessing=None, data=None, **kwargs):
- super(LimeBackend, self).__init__(model, preprocessing)
+ super().__init__(model, preprocessing)
self.explainer = None
self.data = data
@@ -36,11 +37,7 @@ def run_explainer(self, x: pd.DataFrame):
dict containing local contributions
"""
data = self.data if self.data is not None else x
- explainer = lime_tabular.LimeTabularExplainer(
- data.values,
- feature_names=x.columns,
- mode=self._case
- )
+ explainer = lime_tabular.LimeTabularExplainer(data.values, feature_names=x.columns, mode=self._case)
lime_contrib = []
for i in x.index:
@@ -49,8 +46,7 @@ def run_explainer(self, x: pd.DataFrame):
if num_classes <= 2:
exp = explainer.explain_instance(x.loc[i], self.model.predict_proba, num_features=x.shape[1])
- lime_contrib.append(
- dict([[_transform_name(var_name[0], x), var_name[1]] for var_name in exp.as_list()]))
+ lime_contrib.append({_transform_name(var_name[0], x): var_name[1] for var_name in exp.as_list()})
elif num_classes > 2:
contribution = []
@@ -59,11 +55,11 @@ def run_explainer(self, x: pd.DataFrame):
df_contrib = pd.DataFrame()
for i in x.index:
exp = explainer.explain_instance(
- x.loc[i], self.model.predict_proba, top_labels=num_classes,
- num_features=x.shape[1])
+ x.loc[i], self.model.predict_proba, top_labels=num_classes, num_features=x.shape[1]
+ )
list_contrib.append(
- dict([[_transform_name(var_name[0], x), var_name[1]] for var_name in
- exp.as_list(j)]))
+ {_transform_name(var_name[0], x): var_name[1] for var_name in exp.as_list(j)}
+ )
df_contrib = pd.DataFrame(list_contrib)
df_contrib = df_contrib[list(x.columns)]
contribution.append(df_contrib.values)
@@ -71,8 +67,7 @@ def run_explainer(self, x: pd.DataFrame):
else:
exp = explainer.explain_instance(x.loc[i], self.model.predict, num_features=x.shape[1])
- lime_contrib.append(
- dict([[_transform_name(var_name[0], x), var_name[1]] for var_name in exp.as_list()]))
+ lime_contrib.append({_transform_name(var_name[0], x): var_name[1] for var_name in exp.as_list()})
contributions = pd.DataFrame(lime_contrib, index=x.index)
contributions = contributions[list(x.columns)]
@@ -83,9 +78,8 @@ def run_explainer(self, x: pd.DataFrame):
def _transform_name(var_name, x_df):
- """Function for transform name of LIME contribution shape to a comprehensive name
- """
+ """Function for transform name of LIME contribution shape to a comprehensive name"""
for colname in list(x_df.columns):
- if f' {colname} ' in f' {var_name} ':
+ if f" {colname} " in f" {var_name} ":
col_rename = colname
return col_rename
diff --git a/shapash/backend/shap_backend.py b/shapash/backend/shap_backend.py
index 4771758c..716a304a 100644
--- a/shapash/backend/shap_backend.py
+++ b/shapash/backend/shap_backend.py
@@ -1,5 +1,5 @@
-import pandas as pd
import numpy as np
+import pandas as pd
import shap
from shapash.backend.base_backend import BaseBackend
@@ -8,18 +8,18 @@
class ShapBackend(BaseBackend):
# When grouping features contributions together, Shap uses the sum of the contributions
# of the features that belong to the group
- column_aggregation = 'sum'
- name = 'shap'
+ column_aggregation = "sum"
+ name = "shap"
def __init__(self, model, preprocessing=None, masker=None, explainer_args=None, explainer_compute_args=None):
- super(ShapBackend, self).__init__(model, preprocessing)
+ super().__init__(model, preprocessing)
self.masker = masker
self.explainer_args = explainer_args if explainer_args else {}
self.explainer_compute_args = explainer_compute_args if explainer_compute_args else {}
if self.explainer_args:
if "explainer" in self.explainer_args.keys():
- shap_parameters = {k: v for k, v in self.explainer_args.items() if k != 'explainer'}
+ shap_parameters = {k: v for k, v in self.explainer_args.items() if k != "explainer"}
self.explainer = self.explainer_args["explainer"](**shap_parameters)
else:
self.explainer = shap.Explainer(**self.explainer_args)
@@ -31,15 +31,14 @@ def __init__(self, model, preprocessing=None, masker=None, explainer_args=None,
elif shap.explainers.Additive.supports_model_with_masker(model, self.masker):
self.explainer = shap.Explainer(model=model, masker=self.masker)
# otherwise use a model agnostic method
- elif hasattr(model, 'predict_proba'):
+ elif hasattr(model, "predict_proba"):
self.explainer = shap.Explainer(model=model.predict_proba, masker=self.masker)
- elif hasattr(model, 'predict'):
+ elif hasattr(model, "predict"):
self.explainer = shap.Explainer(model=model.predict, masker=self.masker)
# if we get here then we don't know how to handle what was given to us
else:
raise ValueError("The model is not recognized by Shapash! Model: " + str(model))
-
def run_explainer(self, x: pd.DataFrame) -> dict:
"""
Computes and returns local contributions using Shap explainer
@@ -78,8 +77,10 @@ def get_shap_interaction_values(x_df, explainer):
Shap interaction values for each sample as an array of shape (# samples x # features x # features).
"""
if not isinstance(explainer, shap.TreeExplainer):
- raise ValueError(f"Explainer type ({type(explainer)}) is not a TreeExplainer. "
- f"Shap interaction values can only be computed for TreeExplainer types")
+ raise ValueError(
+ f"Explainer type ({type(explainer)}) is not a TreeExplainer. "
+ f"Shap interaction values can only be computed for TreeExplainer types"
+ )
shap_interaction_values = explainer.shap_interaction_values(x_df)
diff --git a/shapash/data/data_loader.py b/shapash/data/data_loader.py
index bf91ca20..e1ad6699 100644
--- a/shapash/data/data_loader.py
+++ b/shapash/data/data_loader.py
@@ -1,12 +1,13 @@
"""
Data loader module
"""
+import json
import os
from pathlib import Path
-import json
-import pandas as pd
-from urllib.request import urlopen
from urllib.error import URLError
+from urllib.request import urlopen
+
+import pandas as pd
def _find_file(data_path, github_data_url, filename):
@@ -29,7 +30,7 @@ def _find_file(data_path, github_data_url, filename):
"""
file = os.path.join(data_path, filename)
if os.path.isfile(file) is False:
- file = github_data_url+filename
+ file = github_data_url + filename
try:
urlopen(file)
except URLError:
@@ -63,42 +64,42 @@ def data_loading(dataset):
If exist, columns labels dictionnary associated to the dataset.
"""
data_path = str(Path(__file__).parents[2] / "data")
- if dataset == 'house_prices':
- github_data_url = 'https://github.com/MAIF/shapash/raw/master/data/'
+ if dataset == "house_prices":
+ github_data_url = "https://github.com/MAIF/shapash/raw/master/data/"
data_house_prices_path = _find_file(data_path, github_data_url, "house_prices_dataset.csv")
dict_house_prices_path = _find_file(data_path, github_data_url, "house_prices_labels.json")
- data = pd.read_csv(data_house_prices_path, header=0, index_col=0, engine='python')
+ data = pd.read_csv(data_house_prices_path, header=0, index_col=0, engine="python")
if github_data_url in dict_house_prices_path:
with urlopen(dict_house_prices_path) as openfile:
dic = json.load(openfile)
else:
- with open(dict_house_prices_path, 'r') as openfile:
+ with open(dict_house_prices_path) as openfile:
dic = json.load(openfile)
return data, dic
- elif dataset == 'titanic':
- github_data_url = 'https://github.com/MAIF/shapash/raw/master/data/'
+ elif dataset == "titanic":
+ github_data_url = "https://github.com/MAIF/shapash/raw/master/data/"
data_titanic_path = _find_file(data_path, github_data_url, "titanicdata.csv")
- dict_titanic_path = _find_file(data_path, github_data_url, 'titaniclabels.json')
- data = pd.read_csv(data_titanic_path, header=0, index_col=0, engine='python')
+ dict_titanic_path = _find_file(data_path, github_data_url, "titaniclabels.json")
+ data = pd.read_csv(data_titanic_path, header=0, index_col=0, engine="python")
if github_data_url in dict_titanic_path:
with urlopen(dict_titanic_path) as openfile:
dic = json.load(openfile)
else:
- with open(dict_titanic_path, 'r') as openfile:
+ with open(dict_titanic_path) as openfile:
dic = json.load(openfile)
return data, dic
- elif dataset == 'telco_customer_churn':
- github_data_url = 'https://github.com/IBM/telco-customer-churn-on-icp4d/raw/master/data/'
+ elif dataset == "telco_customer_churn":
+ github_data_url = "https://github.com/IBM/telco-customer-churn-on-icp4d/raw/master/data/"
data_telco_path = _find_file(data_path, github_data_url, "Telco-Customer-Churn.csv")
- data = pd.read_csv(data_telco_path, header=0, index_col=0, engine='python')
+ data = pd.read_csv(data_telco_path, header=0, index_col=0, engine="python")
return data
- elif dataset == 'us_car_accident':
- github_data_url = 'https://github.com/MAIF/shapash/raw/master/data/'
+ elif dataset == "us_car_accident":
+ github_data_url = "https://github.com/MAIF/shapash/raw/master/data/"
data_accidents_path = _find_file(data_path, github_data_url, "US_Accidents_extract.csv")
- data = pd.read_csv(data_accidents_path, header=0, engine='python')
+ data = pd.read_csv(data_accidents_path, header=0, engine="python")
return data
else:
diff --git a/shapash/decomposition/contributions.py b/shapash/decomposition/contributions.py
index 9820de2f..79f21409 100644
--- a/shapash/decomposition/contributions.py
+++ b/shapash/decomposition/contributions.py
@@ -2,14 +2,15 @@
Contributions
"""
-import pandas as pd
import numpy as np
-from shapash.utils.transform import preprocessing_tolist
-from shapash.utils.transform import check_transformers
+import pandas as pd
+
from shapash.utils.category_encoder_backend import calc_inv_contrib_ce
from shapash.utils.columntransformer_backend import calc_inv_contrib_ct
+from shapash.utils.transform import check_transformers, preprocessing_tolist
-def inverse_transform_contributions(contributions, preprocessing=None, agg_columns='sum'):
+
+def inverse_transform_contributions(contributions, preprocessing=None, agg_columns="sum"):
"""
Reverse contribution giving a preprocessing.
@@ -38,12 +39,12 @@ def inverse_transform_contributions(contributions, preprocessing=None, agg_colum
"""
if not isinstance(contributions, pd.DataFrame):
- raise Exception('Shap values must be a pandas dataframe.')
+ raise Exception("Shap values must be a pandas dataframe.")
if preprocessing is None:
return contributions
else:
- #Transform preprocessing into a list
+ # Transform preprocessing into a list
list_encoding = preprocessing_tolist(preprocessing)
# check supported inverse
@@ -59,6 +60,7 @@ def inverse_transform_contributions(contributions, preprocessing=None, agg_colum
x_contrib_invers = calc_inv_contrib_ce(x_contrib_invers, encoding, agg_columns)
return x_contrib_invers
+
def rank_contributions(s_df, x_df):
"""
Function to sort contributions and input features
@@ -85,14 +87,15 @@ def rank_contributions(s_df, x_df):
sorted_contrib = np.take_along_axis(s_df.values, argsort, axis=1)
sorted_features = np.take_along_axis(x_df.values, argsort, axis=1)
- contrib_col = ['contribution_' + str(i) for i in range(s_df.shape[1])]
- col = ['feature_' + str(i) for i in range(s_df.shape[1])]
+ contrib_col = ["contribution_" + str(i) for i in range(s_df.shape[1])]
+ col = ["feature_" + str(i) for i in range(s_df.shape[1])]
s_dict = pd.DataFrame(data=argsort, columns=col, index=x_df.index)
s_ord = pd.DataFrame(data=sorted_contrib, columns=contrib_col, index=x_df.index)
x_ord = pd.DataFrame(data=sorted_features, columns=col, index=x_df.index)
return [s_ord, x_ord, s_dict]
+
def assign_contributions(ranked):
"""
Turn a list of results into a dict.
@@ -114,11 +117,7 @@ def assign_contributions(ranked):
"""
if len(ranked) != 3:
raise ValueError(
- 'Expected lenght : 3, observed lenght : {},'
- 'please check the outputs of rank_contributions.'.format(len(ranked))
+ "Expected lenght : 3, observed lenght : {},"
+ "please check the outputs of rank_contributions.".format(len(ranked))
)
- return {
- 'contrib_sorted': ranked[0],
- 'x_sorted': ranked[1],
- 'var_dict': ranked[2]
- }
+ return {"contrib_sorted": ranked[0], "x_sorted": ranked[1], "var_dict": ranked[2]}
diff --git a/shapash/explainer/consistency.py b/shapash/explainer/consistency.py
index 0dd56cb3..631287c9 100644
--- a/shapash/explainer/consistency.py
+++ b/shapash/explainer/consistency.py
@@ -1,20 +1,20 @@
-from category_encoders import OrdinalEncoder
import copy
import itertools
+
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
+from category_encoders import OrdinalEncoder
from plotly import graph_objs as go
from plotly.offline import plot
from plotly.subplots import make_subplots
from sklearn.manifold import MDS
from shapash import SmartExplainer
-from shapash.style.style_utils import colors_loading, select_palette, define_style
-
+from shapash.style.style_utils import colors_loading, define_style, select_palette
-class Consistency():
+class Consistency:
def __init__(self):
self._palette_name = list(colors_loading().keys())[0]
self._style_dict = define_style(select_palette(colors_loading(), self._palette_name))
@@ -27,9 +27,8 @@ def tuning_colorscale(self, values):
values whose quantiles must be calculated
"""
desc_df = values.describe(percentiles=np.arange(0.1, 1, 0.1).tolist())
- min_pred, max_init = list(desc_df.loc[['min', 'max']].values)
- desc_pct_df = (desc_df.loc[~desc_df.index.isin(['count', 'mean', 'std'])] - min_pred) / \
- (max_init - min_pred)
+ min_pred, max_init = list(desc_df.loc[["min", "max"]].values)
+ desc_pct_df = (desc_df.loc[~desc_df.index.isin(["count", "mean", "std"])] - min_pred) / (max_init - min_pred)
color_scale = list(map(list, (zip(desc_pct_df.values.flatten(), self._style_dict["init_contrib_colorscale"]))))
return color_scale
@@ -59,7 +58,7 @@ def compile(self, contributions, x=None, preprocessing=None):
self.x = x
self.preprocessing = preprocessing
if not isinstance(contributions, dict):
- raise ValueError('Contributions must be a dictionary')
+ raise ValueError("Contributions must be a dictionary")
self.methods = list(contributions.keys())
self.weights = list(contributions.values())
@@ -77,15 +76,15 @@ def check_consistency_contributions(self, weights):
List of contributions from different methods
"""
if weights[0].ndim == 1:
- raise ValueError('Multiple datapoints are required to compute the metric')
+ raise ValueError("Multiple datapoints are required to compute the metric")
if not all(isinstance(x, pd.DataFrame) for x in weights):
- raise ValueError('Contributions must be pandas DataFrames')
+ raise ValueError("Contributions must be pandas DataFrames")
if not all(x.shape == weights[0].shape for x in weights):
- raise ValueError('Contributions must be of same shape')
+ raise ValueError("Contributions must be of same shape")
if not all(x.columns.tolist() == weights[0].columns.tolist() for x in weights):
- raise ValueError('Columns names are different between contributions')
+ raise ValueError("Columns names are different between contributions")
if not all(x.index.tolist() == weights[0].index.tolist() for x in weights):
- raise ValueError('Index names are different between contributions')
+ raise ValueError("Index names are different between contributions")
def consistency_plot(self, selection=None, max_features=20):
"""
@@ -113,16 +112,17 @@ def consistency_plot(self, selection=None, max_features=20):
weights = [weight.values for weight in self.weights]
elif isinstance(selection, list):
if len(selection) == 1:
- raise ValueError('Selection must include multiple points')
+ raise ValueError("Selection must include multiple points")
else:
weights = [weight.values[selection] for weight in self.weights]
else:
- raise ValueError('Parameter selection must be a list')
+ raise ValueError("Parameter selection must be a list")
all_comparisons, mean_distances = self.calculate_all_distances(self.methods, weights)
- method_1, method_2, l2, index, backend_name_1, backend_name_2 = \
- self.find_examples(mean_distances, all_comparisons, weights)
+ method_1, method_2, l2, index, backend_name_1, backend_name_2 = self.find_examples(
+ mean_distances, all_comparisons, weights
+ )
self.plot_comparison(mean_distances)
self.plot_examples(method_1, method_2, l2, index, backend_name_1, backend_name_2, max_features)
@@ -156,7 +156,12 @@ def calculate_all_distances(self, methods, weights):
l2_dist = self.calculate_pairwise_distances(weights, index_i, index_j)
# Populate the (n choose 2)x4 array
pairwise_comparison = np.column_stack(
- (np.repeat(index_i, len(l2_dist)), np.repeat(index_j, len(l2_dist)), np.arange(len(l2_dist)), l2_dist,)
+ (
+ np.repeat(index_i, len(l2_dist)),
+ np.repeat(index_j, len(l2_dist)),
+ np.arange(len(l2_dist)),
+ l2_dist,
+ )
)
all_comparisons = np.concatenate((all_comparisons, pairwise_comparison), axis=0)
@@ -249,7 +254,9 @@ def find_examples(self, mean_distances, all_comparisons, weights):
l2 = []
# Evenly split the scale of L2 distances (from min to max excluding 0)
- for i in np.linspace(start=mean_distances[mean_distances > 0].min().min(), stop=mean_distances.max().max(), num=5):
+ for i in np.linspace(
+ start=mean_distances[mean_distances > 0].min().min(), stop=mean_distances.max().max(), num=5
+ ):
# For each split, find the closest existing L2 distance
closest_l2 = all_comparisons[:, -1][np.abs(all_comparisons[:, -1] - i).argmin()]
# Return the row that contains this L2 distance
@@ -298,18 +305,25 @@ def plot_comparison(self, mean_distances):
mean_distances : DataFrame
DataFrame storing all pairwise distances between methods
"""
- font = {"color": '#{:02x}{:02x}{:02x}'.format(50, 50, 50)}
+ font = {"color": "#{:02x}{:02x}{:02x}".format(50, 50, 50)}
fig, ax = plt.subplots(ncols=1, figsize=(10, 6))
- ax.text(x=0.5, y=1.04, s="Consistency of explanations:", fontsize=24, ha="center", transform=fig.transFigure, **font)
- ax.text(x=0.5, y=0.98, s="How similar are explanations from different methods?",
- fontsize=18, ha="center", transform=fig.transFigure, **font)
-
- ax.set_title(
- "Average distances between the explanations", fontsize=14, pad=-60
+ ax.text(
+ x=0.5, y=1.04, s="Consistency of explanations:", fontsize=24, ha="center", transform=fig.transFigure, **font
+ )
+ ax.text(
+ x=0.5,
+ y=0.98,
+ s="How similar are explanations from different methods?",
+ fontsize=18,
+ ha="center",
+ transform=fig.transFigure,
+ **font,
)
+ ax.set_title("Average distances between the explanations", fontsize=14, pad=-60)
+
coords = self.calculate_coords(mean_distances)
ax.scatter(coords[:, 0], coords[:, 1], marker="o")
@@ -331,9 +345,9 @@ def plot_comparison(self, mean_distances):
)
# set gray background
- ax.set_facecolor('#F5F5F2')
+ ax.set_facecolor("#F5F5F2")
# draw solid white grid lines
- ax.grid(color='w', linestyle='solid')
+ ax.grid(color="w", linestyle="solid")
lim = (coords.min().min(), coords.max().max())
margin = 0.1 * (lim[1] - lim[0])
@@ -395,8 +409,8 @@ def plot_examples(self, method_1, method_2, l2, index, backend_name_1, backend_n
figure
"""
y = np.arange(method_1[0].shape[0])
- fig, axes = plt.subplots(ncols=len(l2), figsize=(3*len(l2), 4))
- fig.subplots_adjust(wspace=.3, top=.8)
+ fig, axes = plt.subplots(ncols=len(l2), figsize=(3 * len(l2), 4))
+ fig.subplots_adjust(wspace=0.3, top=0.8)
if len(l2) == 1:
axes = np.array([axes])
fig.suptitle("Examples of explanations' comparisons for various distances (L2 norm)")
@@ -411,27 +425,36 @@ def plot_examples(self, method_1, method_2, l2, index, backend_name_1, backend_n
idx = np.flip(i.argsort())
i, j = i[idx], j[idx]
- axes[n].barh(y, i, label='method 1', left=0, color='#{:02x}{:02x}{:02x}'.format(255, 166, 17))
- axes[n].barh(y, j, label='method 2', left=np.abs(np.max(i)) + np.abs(np.min(j)) + np.max(i)/3,
- color='#{:02x}{:02x}{:02x}'.format(117, 152, 189)) # /3 to add space
+ axes[n].barh(y, i, label="method 1", left=0, color="#{:02x}{:02x}{:02x}".format(255, 166, 17))
+ axes[n].barh(
+ y,
+ j,
+ label="method 2",
+ left=np.abs(np.max(i)) + np.abs(np.min(j)) + np.max(i) / 3,
+ color="#{:02x}{:02x}{:02x}".format(117, 152, 189),
+ ) # /3 to add space
# set gray background
- axes[n].set_facecolor('#F5F5F2')
+ axes[n].set_facecolor("#F5F5F2")
# draw solid white grid lines
- axes[n].grid(color='w', linestyle='solid')
+ axes[n].grid(color="w", linestyle="solid")
- axes[n].set(title="%s: %s" %
- (self.index.name if self.index.name is not None else "Id", l) + "\n$d_{L2}$ = " + str(round(k, 2)))
+ axes[n].set(
+ title="{}: {}".format(self.index.name if self.index.name is not None else "Id", l)
+ + "\n$d_{L2}$ = "
+ + str(round(k, 2))
+ )
axes[n].set_xlabel("Contributions")
axes[n].set_ylabel(f"Top {max_features} features")
- axes[n].set_xticks([0, np.abs(np.max(i)) + np.abs(np.min(j)) + np.max(i)/3])
+ axes[n].set_xticks([0, np.abs(np.max(i)) + np.abs(np.min(j)) + np.max(i) / 3])
axes[n].set_xticklabels([m, o])
axes[n].set_yticks([])
return fig
- def pairwise_consistency_plot(self, methods, selection=None,
- max_features=10, max_points=100, file_name=None, auto_open=False):
+ def pairwise_consistency_plot(
+ self, methods, selection=None, max_features=10, max_points=100, file_name=None, auto_open=False
+ ):
"""The Pairwise_Consistency_plot compares the difference of 2 explainability methods across each feature and each data point,
and plots the distribution of those differences.
@@ -463,11 +486,11 @@ def pairwise_consistency_plot(self, methods, selection=None,
figure
"""
if self.x is None:
- raise ValueError('x must be defined in the compile to display the plot')
+ raise ValueError("x must be defined in the compile to display the plot")
if not isinstance(self.x, pd.DataFrame):
- raise ValueError('x must be a pandas DataFrame')
+ raise ValueError("x must be a pandas DataFrame")
if len(methods) != 2:
- raise ValueError('Choose 2 methods among methods of the contributions')
+ raise ValueError("Choose 2 methods among methods of the contributions")
# Select contributions of input methods
pair_indices = [self.methods.index(x) for x in methods]
@@ -480,12 +503,12 @@ def pairwise_consistency_plot(self, methods, selection=None,
x = self.x.iloc[ind_max_points]
elif isinstance(selection, list):
if len(selection) == 1:
- raise ValueError('Selection must include multiple points')
+ raise ValueError("Selection must include multiple points")
else:
weights = [weight.iloc[selection] for weight in pair_weights]
x = self.x.iloc[selection]
else:
- raise ValueError('Parameter selection must be a list')
+ raise ValueError("Parameter selection must be a list")
# Remove constant columns
const_cols = x.loc[:, x.apply(pd.Series.nunique) == 1]
@@ -526,15 +549,17 @@ def plot_pairwise_consistency(self, weights, x, top_features, methods, file_name
if isinstance(self.preprocessing, OrdinalEncoder):
encoder = self.preprocessing
else:
- categorical_features = [col for col in x.columns if x[col].dtype == 'object']
- encoder = OrdinalEncoder(cols=categorical_features,
- handle_unknown='ignore',
- return_df=True).fit(x)
+ categorical_features = [col for col in x.columns if x[col].dtype == "object"]
+ encoder = OrdinalEncoder(cols=categorical_features, handle_unknown="ignore", return_df=True).fit(x)
x = encoder.transform(x)
- xaxis_title = "Difference of contributions between the 2 methods" \
- + f"Feature value : {i}{methods[0]} : {j}{methods[1]} : {k}Diff : {l}"
- for i, j, k, l in zip(feature_value if switch else x[c].round(3),
- weights[0][c].round(2),
- weights[1][c].round(2),
- (weights[0][c] - weights[1][c]).round(2))]
+ hv_text = [
+ f"Feature value : {i}{methods[0]} : {j}{methods[1]} : {k}Diff : {l}"
+ for i, j, k, l in zip(
+ feature_value if switch else x[c].round(3),
+ weights[0][c].round(2),
+ weights[1][c].round(2),
+ (weights[0][c] - weights[1][c]).round(2),
+ )
+ ]
fig.add_trace(
go.Violin(
@@ -565,25 +594,23 @@ def plot_pairwise_consistency(self, weights, x, top_features, methods, file_name
fillcolor="rgba(255, 0, 0, 0.1)",
line={"color": "black", "width": 0.5},
showlegend=False,
- ), secondary_y=False
+ ),
+ secondary_y=False,
)
fig.add_trace(
go.Scatter(
x=(weights[0][c] - weights[1][c]).values,
- y=len(x)*[i] + np.random.normal(0, 0.1, len(x)),
- mode='markers',
- marker={"color": x[c].values,
- "colorscale": self.tuning_colorscale(x[c]),
- "opacity": 0.7},
+ y=len(x) * [i] + np.random.normal(0, 0.1, len(x)),
+ mode="markers",
+ marker={"color": x[c].values, "colorscale": self.tuning_colorscale(x[c]), "opacity": 0.7},
name=c,
- text=len(x)*[c],
+ text=len(x) * [c],
hovertext=hv_text,
- hovertemplate="%{text} %{text} {truncate_str(feature_name)} - Feature Contribution"
# Add subtitle and / or addnote
if subtitle or addnote:
- #title += f"" + subtitle + " - " + addnote + " "
elif subtitle:
@@ -144,47 +149,43 @@ def _update_contributions_fig(self,
dict_t = copy.deepcopy(self._style_dict["dict_title"])
dict_xaxis = copy.deepcopy(self._style_dict["dict_xaxis"])
dict_yaxis = copy.deepcopy(self._style_dict["dict_yaxis"])
- dict_t['text'] = title
- dict_xaxis['text'] = truncate_str(feature_name, 110)
- dict_yaxis['text'] = 'Contribution'
+ dict_t["text"] = title
+ dict_xaxis["text"] = truncate_str(feature_name, 110)
+ dict_yaxis["text"] = "Contribution"
if self.explainer._case == "regression":
colorpoints = pred
- colorbar_title = 'Predicted'
+ colorbar_title = "Predicted"
elif self.explainer._case == "classification":
colorpoints = proba_values
- colorbar_title = 'Predicted Proba'
+ colorbar_title = "Predicted Proba"
if colorpoints is not None:
fig.data[-1].marker.color = colorpoints.values.flatten()
- fig.data[-1].marker.coloraxis = 'coloraxis'
+ fig.data[-1].marker.coloraxis = "coloraxis"
fig.layout.coloraxis.colorscale = col_scale
- fig.layout.coloraxis.colorbar = {'title': {'text': colorbar_title}}
-
- elif fig.data[0].type != 'violin':
- if self.explainer._case == 'classification' and pred is not None:
- fig.data[-1].marker.color = pred.iloc[:, 0].apply(lambda
- x: self._style_dict["violin_area_classif"][1] if x == col_modality else
- self._style_dict["violin_area_classif"][0])
+ fig.layout.coloraxis.colorbar = {"title": {"text": colorbar_title}}
+
+ elif fig.data[0].type != "violin":
+ if self.explainer._case == "classification" and pred is not None:
+ fig.data[-1].marker.color = pred.iloc[:, 0].apply(
+ lambda x: self._style_dict["violin_area_classif"][1]
+ if x == col_modality
+ else self._style_dict["violin_area_classif"][0]
+ )
else:
fig.data[-1].marker.color = self._style_dict["violin_default"]
- fig.update_traces(
- marker={
- 'size': 10,
- 'opacity': 0.8,
- 'line': {'width': 0.8, 'color': 'white'}
- }
- )
+ fig.update_traces(marker={"size": 10, "opacity": 0.8, "line": {"width": 0.8, "color": "white"}})
fig.update_layout(
- template='none',
+ template="none",
title=dict_t,
width=width,
height=height,
xaxis_title=dict_xaxis,
yaxis_title=dict_yaxis,
- hovermode='closest'
+ hovermode="closest",
)
fig.update_yaxes(automargin=True)
@@ -192,22 +193,24 @@ def _update_contributions_fig(self,
if file_name:
plot(fig, filename=file_name, auto_open=auto_open)
- def plot_scatter(self,
- feature_values,
- contributions,
- feature_name,
- pred=None,
- proba_values=None,
- col_modality=None,
- col_scale=None,
- metadata=None,
- addnote=None,
- subtitle=None,
- width=900,
- height=600,
- file_name=None,
- auto_open=False,
- zoom=False):
+ def plot_scatter(
+ self,
+ feature_values,
+ contributions,
+ feature_name,
+ pred=None,
+ proba_values=None,
+ col_modality=None,
+ col_scale=None,
+ metadata=None,
+ addnote=None,
+ subtitle=None,
+ width=900,
+ height=600,
+ file_name=None,
+ auto_open=False,
+ zoom=False,
+ ):
"""
Scatter plot of one feature contribution across the prediction set.
Parameters
@@ -245,95 +248,101 @@ def plot_scatter(self,
fig = go.Figure()
# add break line to X label if necessary
- max_len_by_row = max([
- round(50 / self.explainer.features_desc[feature_values.columns.values[0]]), 8])
+ max_len_by_row = max([round(50 / self.explainer.features_desc[feature_values.columns.values[0]]), 8])
feature_values.iloc[:, 0] = feature_values.iloc[:, 0].apply(
- add_line_break, args=(max_len_by_row, 120,))
+ add_line_break,
+ args=(
+ max_len_by_row,
+ 120,
+ ),
+ )
if pred is not None:
- hv_text = [f"Id: {x}%{hovertext} %{hovertext} %{hovertext} %{hovertext} %{hovertext} %{hovertext} ", "").replace(" ", ""))
@@ -572,48 +602,42 @@ def plot_features_import(self,
]
layout = go.Layout(
- barmode='group',
- template='none',
+ barmode="group",
+ template="none",
autosize=False,
width=width,
height=height,
title=dict_t,
xaxis_title=dict_xaxis,
yaxis_title=dict_yaxis,
- hovermode='closest',
- margin={
- 'l': 160,
- 'r': 0,
- 't': topmargin,
- 'b': 50
- }
+ hovermode="closest",
+ margin={"l": 160, "r": 0, "t": topmargin, "b": 50},
)
# To change ticktext when the x label size is upper than 30 and zoom is False
if (type(feature_imp1.index[0]) == str) & (not zoom):
# change index to abc...abc if its length is upper than 30
- index_val = [
- y.replace(y[24: len(y)-3], '...') if len(y) > 30 else y for y in feature_imp1.index]
+ index_val = [y.replace(y[24 : len(y) - 3], "...") if len(y) > 30 else y for y in feature_imp1.index]
else:
index_val = feature_imp1.index
bar1 = go.Bar(
x=feature_imp1.round(4),
y=feature_imp1.index,
- orientation='h',
- name='Global',
+ orientation="h",
+ name="Global",
marker=dict_style_bar1,
marker_color=marker_color,
- hovertemplate='Feature: %{customdata}" + subtitle + " "
topmargin += 15
- dict_t['text'] = title
- dict_xaxis['text'] = 'Contribution'
- dict_yaxis['text'] = None
+ dict_t["text"] = title
+ dict_xaxis["text"] = "Contribution"
+ dict_yaxis["text"] = None
layout = go.Layout(
- barmode='group',
- template='none',
+ barmode="group",
+ template="none",
width=width,
height=height,
title=dict_t,
xaxis_title=dict_xaxis,
yaxis_title=dict_yaxis,
- yaxis_type='category',
- hovermode='closest',
- margin={
- 'l': 150,
- 'r': 20,
- 't': topmargin,
- 'b': 70
- }
+ yaxis_type="category",
+ hovermode="closest",
+ margin={"l": 150, "r": 20, "t": topmargin, "b": 70},
)
bars = []
for num, expl in enumerate(list(zip(var_dict, x_val, contrib))):
group_name = None
- if expl[1] == '':
- ylabel = '{} '.format(expl[0])
- hoverlabel = '{} '.format(expl[0])
+ if expl[1] == "":
+ ylabel = "{} ".format(expl[0])
+ hoverlabel = "{} ".format(expl[0])
else:
# If bar is a group of features, hovertext includes the values of the features of the group
# And color changes
- if (self.explainer.features_groups is not None
- and self.explainer.inv_features_dict.get(expl[0]) in self.explainer.features_groups.keys()
- and len(index_value) > 0):
+ if (
+ self.explainer.features_groups is not None
+ and self.explainer.inv_features_dict.get(expl[0]) in self.explainer.features_groups.keys()
+ and len(index_value) > 0
+ ):
group_name = self.explainer.inv_features_dict.get(expl[0])
- feat_groups_values = self.explainer.x_init[self.explainer.features_groups[group_name]]\
- .loc[index_value[0]]
- hoverlabel = '{} : {}'.format(add_line_break(self.explainer.features_dict.get(f_name, f_name),
- 40, maxlen=120),
- add_line_break(f_value, 40, maxlen=160))
- for f_name, f_value in feat_groups_values.to_dict().items()
- ])
+ feat_groups_values = self.explainer.x_init[self.explainer.features_groups[group_name]].loc[
+ index_value[0]
+ ]
+ hoverlabel = "{} : {}".format(
+ add_line_break(self.explainer.features_dict.get(f_name, f_name), 40, maxlen=120),
+ add_line_break(f_value, 40, maxlen=160),
+ )
+ for f_name, f_value in feat_groups_values.to_dict().items()
+ ]
+ )
else:
- hoverlabel = '{} : {} : {} : {} : {} '.format(trunc_new_value))
+ ylabel = f"{trunc_new_value} "
contrib_value = expl[2]
# colors
if contrib_value >= 0:
- color = 1 if expl[1] != '' else 0
+ color = 1 if expl[1] != "" else 0
else:
- color = -1 if expl[1] != '' else -2
+ color = -1 if expl[1] != "" else -2
# If the bar is a group of features we modify the color
if group_name is not None:
- bar_color = self._style_dict["featureimp_groups"][0] if color == 1 else self._style_dict["featureimp_groups"][1]
+ bar_color = (
+ self._style_dict["featureimp_groups"][0]
+ if color == 1
+ else self._style_dict["featureimp_groups"][1]
+ )
else:
- bar_color = dict_local_plot_colors[color]['color']
+ bar_color = dict_local_plot_colors[color]["color"]
barobj = go.Bar(
x=[contrib_value],
y=[ylabel],
customdata=[hoverlabel],
- orientation='h',
+ orientation="h",
marker=dict_local_plot_colors[color],
marker_color=bar_color,
showlegend=False,
- hovertemplate='%{customdata}{label_value} - Proba: {pred:.4f} "
elif self.explainer._case == "regression":
- contrib = data['contrib_sorted']
- x_val = data['x_sorted']
- var_dict = data['var_dict']
+ contrib = data["contrib_sorted"]
+ x_val = data["x_sorted"]
+ var_dict = data["var_dict"]
label_num = None
if show_predict is True:
pred_value = self.local_pred(line[0])
@@ -1085,13 +1116,15 @@ def local_plot(self,
else:
var_dict = [self.explainer.features_dict[self.explainer.columns_dict[x]] for x in var_dict]
if show_masked:
- var_dict, x_val, contrib = self.check_masked_contributions(line, var_dict, x_val, contrib, label=label_num)
+ var_dict, x_val, contrib = self.check_masked_contributions(
+ line, var_dict, x_val, contrib, label=label_num
+ )
# Filtering all negative or positive contrib if specify in mask
exclusion = []
- if hasattr(self.explainer, 'mask_params'):
- if self.explainer.mask_params['positive'] is True:
+ if hasattr(self.explainer, "mask_params"):
+ if self.explainer.mask_params["positive"] is True:
exclusion = np.where(np.array(contrib) < 0)[0].tolist()
- elif self.explainer.mask_params['positive'] is False:
+ elif self.explainer.mask_params["positive"] is False:
exclusion = np.where(np.array(contrib) > 0)[0].tolist()
exclusion.sort(reverse=True)
for expl in exclusion:
@@ -1099,22 +1132,25 @@ def local_plot(self,
del x_val[expl]
del contrib[expl]
- fig = self.plot_bar_chart(line, var_dict, x_val, contrib, yaxis_max_label, subtitle, width, height, file_name,
- auto_open, zoom)
+ fig = self.plot_bar_chart(
+ line, var_dict, x_val, contrib, yaxis_max_label, subtitle, width, height, file_name, auto_open, zoom
+ )
return fig
- def contribution_plot(self,
- col,
- selection=None,
- label=-1,
- violin_maxf=10,
- max_points=2000,
- proba=True,
- width=900,
- height=600,
- file_name=None,
- auto_open=False,
- zoom=False):
+ def contribution_plot(
+ self,
+ col,
+ selection=None,
+ label=-1,
+ violin_maxf=10,
+ max_points=2000,
+ proba=True,
+ width=900,
+ height=600,
+ file_name=None,
+ auto_open=False,
+ zoom=False,
+ ):
"""
contribution_plot method diplays a Plotly scatter or violin plot of a selected feature.
It represents the contribution of the selected feature to the predicted value.
@@ -1166,8 +1202,8 @@ def contribution_plot(self,
label_num, _, label_value = self.explainer.check_label_name(label)
if not isinstance(col, (str, int)):
- raise ValueError('parameter col must be string or int.')
- if hasattr(self.explainer, 'inv_features_dict'):
+ raise ValueError("parameter col must be string or int.")
+ if hasattr(self.explainer, "inv_features_dict"):
col = self.explainer.inv_features_dict.get(col, col)
col_is_group = self.explainer.features_groups and col in self.explainer.features_groups.keys()
@@ -1229,19 +1265,27 @@ def contribution_plot(self,
feature_values = self.explainer.x_init.loc[list_ind, col_name]
if col_is_group:
- feature_values = project_feature_values_1d(feature_values, col, self.explainer.x_init,
- self.explainer.x_encoded, self.explainer.preprocessing,
- features_dict=self.explainer.features_dict)
+ feature_values = project_feature_values_1d(
+ feature_values,
+ col,
+ self.explainer.x_init,
+ self.explainer.x_encoded,
+ self.explainer.preprocessing,
+ features_dict=self.explainer.features_dict,
+ )
contrib = subcontrib.loc[list_ind, col].to_frame()
if self.explainer.features_imp is None:
self.explainer.compute_features_import()
- features_imp = self.explainer.features_imp if isinstance(self.explainer.features_imp, pd.Series) \
+ features_imp = (
+ self.explainer.features_imp
+ if isinstance(self.explainer.features_imp, pd.Series)
else self.explainer.features_imp[0]
- top_features_of_group = features_imp.loc[self.explainer.features_groups[col]] \
- .sort_values(ascending=False)[:4].index # Displaying top 4 features
+ )
+ top_features_of_group = (
+ features_imp.loc[self.explainer.features_groups[col]].sort_values(ascending=False)[:4].index
+ ) # Displaying top 4 features
metadata = {
- self.explainer.features_dict[f_name]: self.explainer.x_init[f_name]
- for f_name in top_features_of_group
+ self.explainer.features_dict[f_name]: self.explainer.x_init[f_name] for f_name in top_features_of_group
}
text_group = "Features values were projected on the x axis using t-SNE"
# if group don't show addnote, if not, it's too long
@@ -1257,11 +1301,11 @@ def contribution_plot(self,
if self.explainer.y_pred is not None:
y_pred = self.explainer.y_pred.loc[list_ind]
# Add labels if exist
- if self.explainer._case == 'classification' and self.explainer.label_dict is not None:
+ if self.explainer._case == "classification" and self.explainer.label_dict is not None:
y_pred = y_pred.applymap(lambda x: self.explainer.label_dict[x])
col_value = self.explainer.label_dict[col_value]
# round predict
- elif self.explainer._case == 'regression':
+ elif self.explainer._case == "regression":
if self.round_digit is None:
self.tuning_round_digit()
y_pred = y_pred.applymap(lambda x: round(x, self.round_digit))
@@ -1270,30 +1314,57 @@ def contribution_plot(self,
# selecting the best plot : Scatter, Violin?
if col_value_count > violin_maxf:
- fig = self.plot_scatter(feature_values, contrib, col_label, y_pred,
- proba_values, col_value, col_scale, metadata,
- addnote, subtitle, width, height, file_name,
- auto_open, zoom)
+ fig = self.plot_scatter(
+ feature_values,
+ contrib,
+ col_label,
+ y_pred,
+ proba_values,
+ col_value,
+ col_scale,
+ metadata,
+ addnote,
+ subtitle,
+ width,
+ height,
+ file_name,
+ auto_open,
+ zoom,
+ )
else:
- fig = self.plot_violin(feature_values, contrib, col_label, y_pred,
- proba_values, col_value, col_scale, addnote,
- subtitle, width, height, file_name, auto_open,
- zoom)
+ fig = self.plot_violin(
+ feature_values,
+ contrib,
+ col_label,
+ y_pred,
+ proba_values,
+ col_value,
+ col_scale,
+ addnote,
+ subtitle,
+ width,
+ height,
+ file_name,
+ auto_open,
+ zoom,
+ )
return fig
- def features_importance(self,
- max_features=20,
- selection=None,
- label=-1,
- group_name=None,
- display_groups=True,
- force=False,
- width=900,
- height=500,
- file_name=None,
- auto_open=False,
- zoom=False):
+ def features_importance(
+ self,
+ max_features=20,
+ selection=None,
+ label=-1,
+ group_name=None,
+ display_groups=True,
+ force=False,
+ width=900,
+ height=500,
+ file_name=None,
+ auto_open=False,
+ zoom=False,
+ ):
"""
features_importance display a plotly features importance plot.
in Multiclass Case, this features_importance focus on a label value.
@@ -1344,14 +1415,16 @@ def features_importance(self,
"""
self.explainer.compute_features_import(force=force)
subtitle = None
- title = 'Features Importance'
+ title = "Features Importance"
display_groups = self.explainer.features_groups is not None and display_groups
if display_groups:
if group_name: # Case where we have groups of features and we want to display only features inside a group
if group_name not in self.explainer.features_groups.keys():
- raise ValueError(f"group_name parameter : {group_name} is not in features_groups keys. "
- f"Possible values are : {list(self.explainer.features_groups.keys())}")
- title += f' - {truncate_str(self.explainer.features_dict.get(group_name), 20)}'
+ raise ValueError(
+ f"group_name parameter : {group_name} is not in features_groups keys. "
+ f"Possible values are : {list(self.explainer.features_groups.keys())}"
+ )
+ title += f" - {truncate_str(self.explainer.features_dict.get(group_name), 20)}"
if isinstance(self.explainer.features_imp, list):
features_importance = [
label_feat_imp.loc[label_feat_imp.index.isin(self.explainer.features_groups[group_name])]
@@ -1375,9 +1448,7 @@ def features_importance(self,
global_feat_imp = features_importance[label_num].tail(max_features)
if selection is not None:
subset_feat_imp = self.explainer.backend.get_global_features_importance(
- contributions=contributions[label_num],
- explain_data=self.explainer.explain_data,
- subset=selection
+ contributions=contributions[label_num], explain_data=self.explainer.explain_data, subset=selection
)
else:
subset_feat_imp = None
@@ -1387,13 +1458,11 @@ def features_importance(self,
global_feat_imp = features_importance.tail(max_features)
if selection is not None:
subset_feat_imp = self.explainer.backend.get_global_features_importance(
- contributions=contributions,
- explain_data=self.explainer.explain_data,
- subset=selection
+ contributions=contributions, explain_data=self.explainer.explain_data, subset=selection
)
else:
subset_feat_imp = None
- addnote = ''
+ addnote = ""
if subset_feat_imp is not None:
subset_feat_imp = subset_feat_imp.reindex(global_feat_imp.index)
subset_feat_imp.index = subset_feat_imp.index.map(self.explainer.features_dict)
@@ -1401,45 +1470,47 @@ def features_importance(self,
raise ValueError("selection argument doesn't return any row")
subset_len = len(selection)
total_len = self.explainer.x_init.shape[0]
- addnote = add_text([addnote,
- f"Subset length: {subset_len} ({int(np.round(100 * subset_len / total_len))}%)"],
- sep=" - ")
+ addnote = add_text(
+ [addnote, f"Subset length: {subset_len} ({int(np.round(100 * subset_len / total_len))}%)"], sep=" - "
+ )
if self.explainer.x_init.shape[1] >= max_features:
- addnote = add_text([addnote,
- f"Total number of features: {int(self.explainer.x_init.shape[1])}"],
- sep=" - ")
+ addnote = add_text([addnote, f"Total number of features: {int(self.explainer.x_init.shape[1])}"], sep=" - ")
global_feat_imp.index = global_feat_imp.index.map(self.explainer.features_dict)
if display_groups:
# Bold font for groups of features
global_feat_imp.index = [
- '' + str(f)
+ "" + str(f)
if self.explainer.inv_features_dict.get(f) in self.explainer.features_groups.keys()
- else str(f) for f in global_feat_imp.index
+ else str(f)
+ for f in global_feat_imp.index
]
if subset_feat_imp is not None:
subset_feat_imp.index = [
- '' + str(f)
+ "" + str(f)
if self.explainer.inv_features_dict.get(f) in self.explainer.features_groups.keys()
- else str(f) for f in subset_feat_imp.index
+ else str(f)
+ for f in subset_feat_imp.index
]
- fig = self.plot_features_import(global_feat_imp, subset_feat_imp,
- title, addnote, subtitle, width,
- height, file_name, auto_open, zoom)
+ fig = self.plot_features_import(
+ global_feat_imp, subset_feat_imp, title, addnote, subtitle, width, height, file_name, auto_open, zoom
+ )
return fig
- def plot_line_comparison(self,
- index,
- feature_values,
- contributions,
- predictions=None,
- dict_features=None,
- subtitle=None,
- width=900,
- height=550,
- file_name=None,
- auto_open=False):
+ def plot_line_comparison(
+ self,
+ index,
+ feature_values,
+ contributions,
+ predictions=None,
+ dict_features=None,
+ subtitle=None,
+ width=900,
+ height=550,
+ file_name=None,
+ auto_open=False,
+ ):
"""
Plotly plot for comparisons. Displays
the contributions of several individuals. One line represents
@@ -1479,39 +1550,36 @@ def plot_line_comparison(self,
dict_yaxis = copy.deepcopy(self._style_dict["dict_yaxis"])
if len(index) == 0:
- warnings.warn('No individuals matched', UserWarning)
- dict_t['text'] = "Compare plot - No Matching Reference Entry "
+ warnings.warn("No individuals matched", UserWarning)
+ dict_t["text"] = "Compare plot - No Matching Reference Entry "
elif len(index) < 2:
- warnings.warn('Comparison needs at least 2 individuals', UserWarning)
- dict_t['text'] = "Compare plot - index : " + ' ; '.join(['' + str(id) + ' ' for id in index])
+ warnings.warn("Comparison needs at least 2 individuals", UserWarning)
+ dict_t["text"] = "Compare plot - index : " + " ; ".join(["" + str(id) + " " for id in index])
else:
- dict_t['text'] = "Compare plot - index : " + ' ; '.join(['' + str(id) + ' ' for id in index])
+ dict_t["text"] = "Compare plot - index : " + " ; ".join(["" + str(id) + " " for id in index])
- dict_xaxis['text'] = "Contributions"
+ dict_xaxis["text"] = "Contributions"
- dict_yaxis['text'] = None
+ dict_yaxis["text"] = None
if subtitle is not None:
topmargin += 15 * height / 275
- dict_t['text'] = truncate_str(dict_t['text'], 120) \
+ dict_t["text"] = (
+ truncate_str(dict_t["text"], 120)
+ f"' + str(feat) + ' ')
+ features.append("" + str(feat) + " ")
pred_x_val = x_val[dict_features[feat]]
- x_hover.append(f"Id: {add_line_break(id_i, 40, 160)} "
- + f"{add_line_break(feat, 40, 160)} {index[i]} ",
- hoverinfo="text",
- hovertext=x_hover,
- marker={'color': dic_color[i % len(dic_color)]}
- )
+ x_hover.append(
+ f"Id: {add_line_break(id_i, 40, 160)} "
+ + f"{add_line_break(feat, 40, 160)} {index[i]} ",
+ hoverinfo="text",
+ hovertext=x_hover,
+ marker={"color": dic_color[i % len(dic_color)]},
+ )
)
fig = go.Figure(data=lines, layout=layout)
@@ -1555,16 +1626,18 @@ def plot_line_comparison(self,
return fig
- def compare_plot(self,
- index=None,
- row_num=None,
- label=None,
- max_features=20,
- width=900,
- height=550,
- show_predict=True,
- file_name=None,
- auto_open=True):
+ def compare_plot(
+ self,
+ index=None,
+ row_num=None,
+ label=None,
+ max_features=20,
+ width=900,
+ height=550,
+ show_predict=True,
+ file_name=None,
+ auto_open=True,
+ ):
"""
Plotly comparison plot of several individuals' contributions. Plots contributions feature by feature.
Allows to see the differences of contributions between two or more individuals,
@@ -1603,9 +1676,7 @@ def compare_plot(self,
"""
# Checking input is okay
if sum(arg is not None for arg in [row_num, index]) != 1:
- raise ValueError(
- "You have to specify just one of these arguments: index, row_num"
- )
+ raise ValueError("You have to specify just one of these arguments: index, row_num")
# Getting indexes in a list
line_reference = []
if index is not None:
@@ -1614,16 +1685,18 @@ def compare_plot(self,
line_reference.append(ident)
elif row_num is not None:
- line_reference = [self.explainer.x_init.index.values[row_nb_reference]
- for row_nb_reference in row_num
- if self.explainer.x_init.index.values[row_nb_reference] in self.explainer.x_init.index]
+ line_reference = [
+ self.explainer.x_init.index.values[row_nb_reference]
+ for row_nb_reference in row_num
+ if self.explainer.x_init.index.values[row_nb_reference] in self.explainer.x_init.index
+ ]
subtitle = ""
if len(line_reference) < 1:
- raise ValueError('No matching entry for index')
+ raise ValueError("No matching entry for index")
# Classification case
- if self.explainer._case == 'classification':
+ if self.explainer._case == "classification":
if label is None:
label = -1
@@ -1632,19 +1705,23 @@ def compare_plot(self,
if show_predict:
preds = [self.local_pred(line, label_num) for line in line_reference]
- subtitle = f"Response: {label_value} - " \
- + "Probas: " \
- + ' ; '.join([str(id) + ': ' + str(round(proba, 2)) + ' '
- for proba, id in zip(preds, line_reference)])
+ subtitle = (
+ f"Response: {label_value} - "
+ + "Probas: "
+ + " ; ".join(
+ [str(id) + ": " + str(round(proba, 2)) + " " for proba, id in zip(preds, line_reference)]
+ )
+ )
# Regression case
- elif self.explainer._case == 'regression':
+ elif self.explainer._case == "regression":
contrib = self.explainer.contributions
if show_predict:
preds = [self.local_pred(line) for line in line_reference]
- subtitle = "Predictions: " + ' ; '.join([str(id) + ': ' + str(round(pred, 2)) + ' '
- for id, pred in zip(line_reference, preds)])
+ subtitle = "Predictions: " + " ; ".join(
+ [str(id) + ": " + str(round(pred, 2)) + " " for id, pred in zip(line_reference, preds)]
+ )
new_contrib = list()
for ident in line_reference:
@@ -1653,7 +1730,7 @@ def compare_plot(self,
# Well labels if available
feature_values = [0] * len(contrib.columns)
- if hasattr(self.explainer, 'columns_dict'):
+ if hasattr(self.explainer, "columns_dict"):
for i, name in enumerate(contrib.columns):
feature_name = self.explainer.features_dict[name]
feature_values[i] = feature_name
@@ -1667,21 +1744,22 @@ def compare_plot(self,
iteration_list = iteration_list[::-1]
new_contrib, feature_values = list(zip(*iteration_list))
- fig = self.plot_line_comparison(line_reference, feature_values, new_contrib,
- predictions=preds, dict_features=dict_features,
- width=width, height=height, subtitle=subtitle,
- file_name=file_name, auto_open=auto_open)
+ fig = self.plot_line_comparison(
+ line_reference,
+ feature_values,
+ new_contrib,
+ predictions=preds,
+ dict_features=dict_features,
+ width=width,
+ height=height,
+ subtitle=subtitle,
+ file_name=file_name,
+ auto_open=auto_open,
+ )
return fig
- def _plot_interactions_scatter(self,
- x_name,
- y_name,
- col_name,
- x_values,
- y_values,
- col_values,
- col_scale):
+ def _plot_interactions_scatter(self, x_name, y_name, col_name, x_values, y_values, col_values, col_scale):
"""
Function used to generate a scatter plot figure for the interactions plots.
Parameters
@@ -1706,32 +1784,38 @@ def _plot_interactions_scatter(self,
"""
# add break line to X label if necessary
max_len_by_row = max([round(50 / self.explainer.features_desc[x_values.columns.values[0]]), 8])
- x_values.iloc[:, 0] = x_values.iloc[:, 0].apply(add_line_break, args=(max_len_by_row, 120,))
+ x_values.iloc[:, 0] = x_values.iloc[:, 0].apply(
+ add_line_break,
+ args=(
+ max_len_by_row,
+ 120,
+ ),
+ )
- data_df = pd.DataFrame({
- x_name: x_values.values.flatten(),
- y_name: y_values.values.flatten(),
- col_name: col_values.values.flatten()
- })
+ data_df = pd.DataFrame(
+ {
+ x_name: x_values.values.flatten(),
+ y_name: y_values.values.flatten(),
+ col_name: col_values.values.flatten(),
+ }
+ )
if isinstance(col_values.values.flatten()[0], str):
- fig = px.scatter(data_df, x=x_name, y=y_name, color=col_name,
- color_discrete_sequence=self._style_dict["interactions_discrete_colors"])
+ fig = px.scatter(
+ data_df,
+ x=x_name,
+ y=y_name,
+ color=col_name,
+ color_discrete_sequence=self._style_dict["interactions_discrete_colors"],
+ )
else:
fig = px.scatter(data_df, x=x_name, y=y_name, color=col_name, color_continuous_scale=col_scale)
- fig.update_traces(mode='markers')
+ fig.update_traces(mode="markers")
return fig
- def _plot_interactions_violin(self,
- x_name,
- y_name,
- col_name,
- x_values,
- y_values,
- col_values,
- col_scale):
+ def _plot_interactions_violin(self, x_name, y_name, col_name, x_values, y_values, col_values, col_scale):
"""
Function used to generate a violin plot figure for the interactions plots.
Parameters
@@ -1759,47 +1843,54 @@ def _plot_interactions_violin(self,
# add break line to X label
max_len_by_row = max([round(50 / self.explainer.features_desc[x_values.columns.values[0]]), 8])
- x_values.iloc[:, 0] = x_values.iloc[:, 0].apply(add_line_break, args=(max_len_by_row, 120,))
+ x_values.iloc[:, 0] = x_values.iloc[:, 0].apply(
+ add_line_break,
+ args=(
+ max_len_by_row,
+ 120,
+ ),
+ )
uniq_l = list(pd.unique(x_values.values.flatten()))
uniq_l.sort()
for i in uniq_l:
- fig.add_trace(go.Violin(x=x_values.loc[x_values.iloc[:, 0] == i].values.flatten(),
- y=y_values.loc[x_values.iloc[:, 0] == i].values.flatten(),
- line_color=self._style_dict["violin_default"],
- showlegend=False,
- meanline_visible=True,
- scalemode='count',
- ))
- scatter_fig = self._plot_interactions_scatter(x_name=x_name, y_name=y_name, col_name=col_name,
- x_values=x_values, y_values=y_values, col_values=col_values,
- col_scale=col_scale)
+ fig.add_trace(
+ go.Violin(
+ x=x_values.loc[x_values.iloc[:, 0] == i].values.flatten(),
+ y=y_values.loc[x_values.iloc[:, 0] == i].values.flatten(),
+ line_color=self._style_dict["violin_default"],
+ showlegend=False,
+ meanline_visible=True,
+ scalemode="count",
+ )
+ )
+ scatter_fig = self._plot_interactions_scatter(
+ x_name=x_name,
+ y_name=y_name,
+ col_name=col_name,
+ x_values=x_values,
+ y_values=y_values,
+ col_values=col_values,
+ col_scale=col_scale,
+ )
for trace in scatter_fig.data:
fig.add_trace(trace)
fig.update_layout(
autosize=False,
- hovermode='closest',
+ hovermode="closest",
violingap=0.05,
violingroupgap=0,
- violinmode='overlay',
- xaxis_type='category'
+ violinmode="overlay",
+ xaxis_type="category",
)
fig.update_xaxes(range=[-0.6, len(uniq_l) - 0.4])
return fig
- def _update_interactions_fig(self,
- fig,
- col_name1,
- col_name2,
- addnote,
- width,
- height,
- file_name,
- auto_open):
+ def _update_interactions_fig(self, fig, col_name1, col_name2, addnote, width, height, file_name, auto_open):
"""
Function used for the interactions plot to update the layout of the plotly figure.
Parameters
@@ -1823,7 +1914,7 @@ def _update_interactions_fig(self,
go.Figure
"""
- if fig.data[-1]['showlegend'] is False: # Case where col2 is not categorical
+ if fig.data[-1]["showlegend"] is False: # Case where col2 is not categorical
fig.layout.coloraxis.colorscale = self._style_dict["interactions_col_scale"]
else:
fig.update_layout(legend=dict(title=dict(text=col_name2)))
@@ -1832,30 +1923,24 @@ def _update_interactions_fig(self,
if addnote:
title += f"%{text} %{text} Feature: {col} Feature: {col} {index} "
title += "{label_value} "
# Plot distribution
- fig.add_trace(go.Violin(
- x=df_pred['target'].values.flatten(),
- y=df_pred['proba_values'].values.flatten(),
- points=False,
- legendgroup='M', scalegroup='M', name='Correct Prediction',
- line_color=self._style_dict["violin_area_classif"][1],
- pointpos=-0.1,
- showlegend=False,
- jitter=0.075,
- meanline_visible=True,
- spanmode="hard",
- customdata=df_pred['proba_values'].index.values,
- scalemode='count',
- ))
+ fig.add_trace(
+ go.Violin(
+ x=df_pred["target"].values.flatten(),
+ y=df_pred["proba_values"].values.flatten(),
+ points=False,
+ legendgroup="M",
+ scalegroup="M",
+ name="Correct Prediction",
+ line_color=self._style_dict["violin_area_classif"][1],
+ pointpos=-0.1,
+ showlegend=False,
+ jitter=0.075,
+ meanline_visible=True,
+ spanmode="hard",
+ customdata=df_pred["proba_values"].index.values,
+ scalemode="count",
+ )
+ )
# Plot points depending if wrong or correct prediction
- df_correct_predict = df_pred[(df_pred['wrong_predict'] == 0)]
- df_wrong_predict = df_pred[(df_pred['wrong_predict'] == 1)]
- hv_text_correct_predict = [f"Id: {x}%{hovertext} %{hovertext} %{hovertext} %{hovertext} %{hovertext} %{hovertext} {{ col['name'] }} - {{ col['type'] }}
{% endfor %}
-{% endfor %}
\ No newline at end of file
+{% endfor %}
diff --git a/shapash/report/html/table_two_columns.html b/shapash/report/html/table_two_columns.html
index 8402242e..962641da 100644
--- a/shapash/report/html/table_two_columns.html
+++ b/shapash/report/html/table_two_columns.html
@@ -16,4 +16,4 @@
{% endfor %}
-
\ No newline at end of file
+
diff --git a/shapash/report/html/univariate.html b/shapash/report/html/univariate.html
index 42fc2118..b29da33f 100644
--- a/shapash/report/html/univariate.html
+++ b/shapash/report/html/univariate.html
@@ -15,4 +15,4 @@ {{ col['name'] }} - {{ col['type'] }}
{{ col['image'] }}
-{% endfor %}
\ No newline at end of file
+{% endfor %}
diff --git a/shapash/report/plots.py b/shapash/report/plots.py
index 420dfdd7..e611a33c 100644
--- a/shapash/report/plots.py
+++ b/shapash/report/plots.py
@@ -1,21 +1,18 @@
-from typing import Union, Optional
-import pandas as pd
+from typing import Optional, Union
+
+import matplotlib.pyplot as plt
import numpy as np
+import pandas as pd
import seaborn as sns
-import matplotlib.pyplot as plt
from matplotlib.colors import LinearSegmentedColormap
-from shapash.utils.utils import truncate_str
from shapash.report.common import VarType
-from shapash.style.style_utils import get_pyplot_color, get_palette
+from shapash.style.style_utils import get_palette, get_pyplot_color
+from shapash.utils.utils import truncate_str
def generate_fig_univariate(
- df_all: pd.DataFrame,
- col: str,
- hue: str,
- type: VarType,
- colors_dict: Optional[dict] = None
+ df_all: pd.DataFrame, col: str, hue: str, type: VarType, colors_dict: Optional[dict] = None
) -> plt.Figure:
"""
Returns a matplotlib figure containing the distribution of any kind of feature
@@ -55,10 +52,7 @@ def generate_fig_univariate(
def generate_fig_univariate_continuous(
- df_all: pd.DataFrame,
- col: str,
- hue: str,
- colors_dict: Optional[dict] = None
+ df_all: pd.DataFrame, col: str, hue: str, colors_dict: Optional[dict] = None
) -> plt.Figure:
"""
Returns a matplotlib figure containing the distribution of a continuous feature.
@@ -78,9 +72,16 @@ def generate_fig_univariate_continuous(
-------
matplotlib.pyplot.Figure
"""
- colors_dict = colors_dict or get_palette('default')
- g = sns.displot(df_all, x=col, hue=hue, kind="kde", fill=True, common_norm=False,
- palette=get_pyplot_color(colors=colors_dict['report_feature_distribution']))
+ colors_dict = colors_dict or get_palette("default")
+ g = sns.displot(
+ df_all,
+ x=col,
+ hue=hue,
+ kind="kde",
+ fill=True,
+ common_norm=False,
+ palette=get_pyplot_color(colors=colors_dict["report_feature_distribution"]),
+ )
g.set_xticklabels(rotation=30)
fig = g.fig
@@ -92,11 +93,7 @@ def generate_fig_univariate_continuous(
def generate_fig_univariate_categorical(
- df_all: pd.DataFrame,
- col: str,
- hue: str,
- nb_cat_max: int = 7,
- colors_dict: Optional[dict] = None
+ df_all: pd.DataFrame, col: str, hue: str, nb_cat_max: int = 7, colors_dict: Optional[dict] = None
) -> plt.Figure:
"""
Returns a matplotlib figure containing the distribution of a categorical feature.
@@ -124,40 +121,50 @@ def generate_fig_univariate_categorical(
-------
matplotlib.pyplot.Figure
"""
- colors_dict = colors_dict or get_palette('default')
- df_cat = df_all.groupby([col, hue]).agg({col: 'count'})\
- .rename(columns={col: "count"}).reset_index()
- df_cat['Percent'] = df_cat['count'] * 100 / df_cat.groupby(hue)['count'].transform('sum')
+ colors_dict = colors_dict or get_palette("default")
+ df_cat = df_all.groupby([col, hue]).agg({col: "count"}).rename(columns={col: "count"}).reset_index()
+ df_cat["Percent"] = df_cat["count"] * 100 / df_cat.groupby(hue)["count"].transform("sum")
if pd.api.types.is_numeric_dtype(df_cat[col].dtype):
df_cat = df_cat.sort_values(col, ascending=True)
df_cat[col] = df_cat[col].astype(str)
- nb_cat = df_cat.groupby([col]).agg({'count': 'sum'}).reset_index()[col].nunique()
+ nb_cat = df_cat.groupby([col]).agg({"count": "sum"}).reset_index()[col].nunique()
if nb_cat > nb_cat_max:
df_cat = _merge_small_categories(df_cat=df_cat, col=col, hue=hue, nb_cat_max=nb_cat_max)
fig, ax = plt.subplots(figsize=(7, 4))
- sns.barplot(data=df_cat, x='Percent', y=col, hue=hue,
- palette=get_pyplot_color(colors=colors_dict['report_feature_distribution']), ax=ax)
+ sns.barplot(
+ data=df_cat,
+ x="Percent",
+ y=col,
+ hue=hue,
+ palette=get_pyplot_color(colors=colors_dict["report_feature_distribution"]),
+ ax=ax,
+ )
for p in ax.patches:
- ax.annotate("{:.1f}%".format(np.nan_to_num(p.get_width(), nan=0)),
- xy=(p.get_width(), p.get_y() + p.get_height() / 2),
- xytext=(5, 0), textcoords='offset points', ha="left", va="center")
+ ax.annotate(
+ "{:.1f}%".format(np.nan_to_num(p.get_width(), nan=0)),
+ xy=(p.get_width(), p.get_y() + p.get_height() / 2),
+ xytext=(5, 0),
+ textcoords="offset points",
+ ha="left",
+ va="center",
+ )
# Shrink current axis by 20%
box = ax.get_position()
ax.set_position([box.x0, box.y0, box.width * 0.8, box.height])
# Put a legend to the right of the current axis
- ax.legend(loc='center left', bbox_to_anchor=(1, 0.5))
+ ax.legend(loc="center left", bbox_to_anchor=(1, 0.5))
# Removes plot borders
- ax.spines['top'].set_visible(False)
- ax.spines['right'].set_visible(False)
+ ax.spines["top"].set_visible(False)
+ ax.spines["right"].set_visible(False)
new_labels = [truncate_str(i.get_text(), maxlen=45) for i in ax.yaxis.get_ticklabels()]
ax.yaxis.set_ticklabels(new_labels)
@@ -165,23 +172,22 @@ def generate_fig_univariate_categorical(
return fig
-def _merge_small_categories(df_cat: pd.DataFrame, col: str, hue: str, nb_cat_max: int) -> pd.DataFrame:
+def _merge_small_categories(df_cat: pd.DataFrame, col: str, hue: str, nb_cat_max: int) -> pd.DataFrame:
"""
Merges categories of column 'col' of df_cat into 'Other' category so that
the number of categories is less than nb_cat_max.
"""
- df_cat_sum_hue = df_cat.groupby([col]).agg({'count': 'sum'}).reset_index()
- list_cat_to_merge = df_cat_sum_hue.sort_values('count', ascending=False)[col].to_list()[nb_cat_max - 1:]
- df_cat_other = df_cat.loc[df_cat[col].isin(list_cat_to_merge)] \
- .groupby(hue, as_index=False)[["count", "Percent"]].sum()
+ df_cat_sum_hue = df_cat.groupby([col]).agg({"count": "sum"}).reset_index()
+ list_cat_to_merge = df_cat_sum_hue.sort_values("count", ascending=False)[col].to_list()[nb_cat_max - 1 :]
+ df_cat_other = (
+ df_cat.loc[df_cat[col].isin(list_cat_to_merge)].groupby(hue, as_index=False)[["count", "Percent"]].sum()
+ )
df_cat_other[col] = "Other"
- return pd.concat([df_cat.loc[~df_cat[col].isin(list_cat_to_merge)],df_cat_other], axis=0)
+ return pd.concat([df_cat.loc[~df_cat[col].isin(list_cat_to_merge)], df_cat_other], axis=0)
def generate_confusion_matrix_plot(
- y_true: Union[np.array, list],
- y_pred: Union[np.array, list],
- colors_dict: Optional[dict] = None
+ y_true: Union[np.array, list], y_pred: Union[np.array, list], colors_dict: Optional[dict] = None
) -> plt.Figure:
"""
Returns a matplotlib figure containing a confusion matrix that is computed using y_true and
@@ -199,11 +205,11 @@ def generate_confusion_matrix_plot(
-------
matplotlib.pyplot.Figure
"""
- colors_dict = colors_dict or get_palette('default')
- col_scale = get_pyplot_color(colors=colors_dict['report_confusion_matrix'])
- cmap_gradient = LinearSegmentedColormap.from_list('col_corr', col_scale, N=100)
+ colors_dict = colors_dict or get_palette("default")
+ col_scale = get_pyplot_color(colors=colors_dict["report_confusion_matrix"])
+ cmap_gradient = LinearSegmentedColormap.from_list("col_corr", col_scale, N=100)
- df_cm = pd.crosstab(y_true, y_pred, rownames=['Actual'], colnames=['Predicted'])
+ df_cm = pd.crosstab(y_true, y_pred, rownames=["Actual"], colnames=["Predicted"])
fig, ax = plt.subplots(figsize=(7, 4))
- sns.heatmap(df_cm, ax=ax, annot=True, cmap=cmap_gradient, fmt='g')
+ sns.heatmap(df_cm, ax=ax, annot=True, cmap=cmap_gradient, fmt="g")
return fig
diff --git a/shapash/report/project_report.py b/shapash/report/project_report.py
index 149a71a2..3c578d01 100644
--- a/shapash/report/project_report.py
+++ b/shapash/report/project_report.py
@@ -1,28 +1,34 @@
-from typing import Optional, Union, Tuple
import logging
-import sys
import os
-from numbers import Number
+import sys
from datetime import date
+from numbers import Number
+from typing import Optional, Tuple, Union
+
import jinja2
-import pandas as pd
import numpy as np
+import pandas as pd
import plotly
-from shapash.utils.transform import inverse_transform, apply_postprocessing, handle_categorical_missing
from shapash import SmartExplainer
+from shapash.report.common import VarType, compute_col_types, display_value, get_callable, series_dtype
+from shapash.report.data_analysis import perform_global_dataframe_analysis, perform_univariate_dataframe_analysis
+from shapash.report.plots import generate_confusion_matrix_plot, generate_fig_univariate
+from shapash.report.visualisation import (
+ convert_fig_to_html,
+ print_css_style,
+ print_html,
+ print_javascript_misc,
+ print_md,
+)
from shapash.utils.io import load_yml
+from shapash.utils.transform import apply_postprocessing, handle_categorical_missing, inverse_transform
from shapash.utils.utils import get_project_root, truncate_str
-from shapash.report.visualisation import print_md, print_html, print_css_style, convert_fig_to_html, \
- print_javascript_misc
-from shapash.report.data_analysis import perform_global_dataframe_analysis, perform_univariate_dataframe_analysis
-from shapash.report.plots import generate_fig_univariate, generate_confusion_matrix_plot
-from shapash.report.common import series_dtype, get_callable, compute_col_types, VarType, display_value
from shapash.webapp.utils.utils import round_to_k
logging.basicConfig(level=logging.INFO)
-template_loader = jinja2.FileSystemLoader(searchpath=os.path.join(get_project_root(), 'shapash', 'report', 'html'))
+template_loader = jinja2.FileSystemLoader(searchpath=os.path.join(get_project_root(), "shapash", "report", "html"))
template_env = jinja2.Environment(loader=template_loader)
@@ -60,13 +66,13 @@ class ProjectReport:
"""
def __init__(
- self,
- explainer: SmartExplainer,
- project_info_file: str,
- x_train: Optional[pd.DataFrame] = None,
- y_train: Optional[pd.DataFrame] = None,
- y_test: Optional[pd.DataFrame] = None,
- config: Optional[dict] = None
+ self,
+ explainer: SmartExplainer,
+ project_info_file: str,
+ x_train: Optional[pd.DataFrame] = None,
+ y_train: Optional[pd.DataFrame] = None,
+ y_test: Optional[pd.DataFrame] = None,
+ config: Optional[dict] = None,
):
self.explainer = explainer
self.metadata = load_yml(path=project_info_file)
@@ -87,35 +93,34 @@ def __init__(
self.y_pred = np.array(self.explainer.y_pred.T)[0]
else:
self.y_pred = self.explainer.model.predict(self.explainer.x_encoded)
- self.y_test, target_name_test = self._get_values_and_name(y_test, 'target')
- self.y_train, target_name_train = self._get_values_and_name(y_train, 'target')
+ self.y_test, target_name_test = self._get_values_and_name(y_test, "target")
+ self.y_train, target_name_train = self._get_values_and_name(y_train, "target")
self.target_name = target_name_train or target_name_test
- if 'title_story' in self.config.keys():
- self.title_story = config['title_story']
- elif self.explainer.title_story != '':
+ if "title_story" in self.config.keys():
+ self.title_story = config["title_story"]
+ elif self.explainer.title_story != "":
self.title_story = self.explainer.title_story
else:
- self.title_story = 'Shapash report'
- self.title_description = self.config['title_description'] if 'title_description' in self.config.keys() else ''
+ self.title_story = "Shapash report"
+ self.title_description = self.config["title_description"] if "title_description" in self.config.keys() else ""
print_css_style()
print_javascript_misc()
- if 'metrics' in self.config.keys():
- if not isinstance(self.config['metrics'], list) or not isinstance(self.config['metrics'][0], dict):
+ if "metrics" in self.config.keys():
+ if not isinstance(self.config["metrics"], list) or not isinstance(self.config["metrics"][0], dict):
raise ValueError("The metrics parameter expects a list of dict.")
- for metric in self.config['metrics']:
+ for metric in self.config["metrics"]:
for key in metric:
- if key not in ['path', 'name', 'use_proba_values']:
+ if key not in ["path", "name", "use_proba_values"]:
raise ValueError(f"Unknown key : {key}. Key should be in ['path', 'name', 'use_proba_values']")
- if key == 'use_proba_values' and not isinstance(metric['use_proba_values'], bool):
+ if key == "use_proba_values" and not isinstance(metric["use_proba_values"], bool):
raise ValueError('"use_proba_values" metric key expects a boolean value.')
@staticmethod
def _get_values_and_name(
- y: Optional[Union[pd.DataFrame, pd.Series, list]],
- default_name: str
+ y: Optional[Union[pd.DataFrame, pd.Series, list]], default_name: str
) -> Union[Tuple[list, str], Tuple[None, None]]:
"""
Extracts vales and column name from a Pandas Series, DataFrame, or assign a default
@@ -170,20 +175,25 @@ def _create_train_test_df(test: Optional[pd.DataFrame], train: Optional[pd.DataF
The concatenation of train and test as a dataframe containing train and test values with
a new 'data_train_test' column allowing to distinguish the values.
"""
- if (test is not None and 'data_train_test' in test.columns) or \
- (train is not None and 'data_train_test' in train.columns):
+ if (test is not None and "data_train_test" in test.columns) or (
+ train is not None and "data_train_test" in train.columns
+ ):
raise ValueError('"data_train_test" column must be renamed as it is used in ProjectReport')
if test is None and train is None:
return None
- return pd.concat([test.assign(data_train_test="test") if test is not None else None,
- train.assign(data_train_test="train") if train is not None else None]).reset_index(drop=True)
+ return pd.concat(
+ [
+ test.assign(data_train_test="test") if test is not None else None,
+ train.assign(data_train_test="train") if train is not None else None,
+ ]
+ ).reset_index(drop=True)
def display_title_description(self):
"""
Displays title of the report and its description if defined.
"""
print_html(f"""{self.title_story} """)
- if self.title_description != '':
+ if self.title_description != "":
print_html(f'{self.title_description} ')
def display_project_information(self):
@@ -193,11 +203,11 @@ def display_project_information(self):
for section in self.metadata.keys():
print_md(f"## {section.title()}")
for k, v in self.metadata[section].items():
- if k.lower() == 'date' and v.lower() == 'auto':
+ if k.lower() == "date" and v.lower() == "auto":
print_md(f"**{k.title()}** : {date.today()}")
else:
print_md(f"**{k.title()}** : {v}")
- print_md('---')
+ print_md("---")
def display_model_analysis(self):
"""
@@ -209,29 +219,37 @@ def display_model_analysis(self):
print_md(f"**Library :** {self.explainer.model.__class__.__module__}")
for name, module in sorted(sys.modules.items()):
- if hasattr(module, '__version__') \
- and self.explainer.model.__class__.__module__.split('.')[0] == module.__name__:
+ if (
+ hasattr(module, "__version__")
+ and self.explainer.model.__class__.__module__.split(".")[0] == module.__name__
+ ):
print_md(f"**Library version :** {module.__version__}")
print_md("**Model parameters :** ")
model_params = self.explainer.model.__dict__
table_template = template_env.get_template("double_table.html")
- print_html(table_template.render(
- columns1=["Parameter key", "Parameter value"],
- rows1=[{"name": truncate_str(str(k), 50), "value": truncate_str(str(v), 300)}
- for k, v in list(model_params.items())[:len(model_params)//2:]], # Getting half of the parameters
- columns2=["Parameter key", "Parameter value"],
- rows2=[{"name": truncate_str(str(k), 50), "value": truncate_str(str(v), 300)}
- for k, v in list(model_params.items())[len(model_params)//2:]] # Getting 2nd half of the parameters
- ))
- print_md('---')
+ print_html(
+ table_template.render(
+ columns1=["Parameter key", "Parameter value"],
+ rows1=[
+ {"name": truncate_str(str(k), 50), "value": truncate_str(str(v), 300)}
+ for k, v in list(model_params.items())[: len(model_params) // 2 :]
+ ], # Getting half of the parameters
+ columns2=["Parameter key", "Parameter value"],
+ rows2=[
+ {"name": truncate_str(str(k), 50), "value": truncate_str(str(v), 300)}
+ for k, v in list(model_params.items())[len(model_params) // 2 :]
+ ], # Getting 2nd half of the parameters
+ )
+ )
+ print_md("---")
def display_dataset_analysis(
- self,
- global_analysis: bool = True,
- univariate_analysis: bool = True,
- target_analysis: bool = True,
- multivariate_analysis: bool = True
+ self,
+ global_analysis: bool = True,
+ univariate_analysis: bool = True,
+ target_analysis: bool = True,
+ multivariate_analysis: bool = True,
):
"""
This method performs and displays an exploration of the data given.
@@ -262,14 +280,16 @@ def display_dataset_analysis(
col_splitter="data_train_test",
split_values=["test", "train"],
names=["Prediction dataset", "Training dataset"],
- group_id='univariate'
+ group_id="univariate",
)
if target_analysis:
df_target = self._create_train_test_df(
- test=pd.DataFrame({self.target_name: self.y_test},
- index=range(len(self.y_test))) if self.y_test is not None else None,
- train=pd.DataFrame({self.target_name: self.y_train},
- index=range(len(self.y_train))) if self.y_train is not None else None
+ test=pd.DataFrame({self.target_name: self.y_test}, index=range(len(self.y_test)))
+ if self.y_test is not None
+ else None,
+ train=pd.DataFrame({self.target_name: self.y_train}, index=range(len(self.y_train)))
+ if self.y_train is not None
+ else None,
)
if df_target is not None:
if target_analysis:
@@ -279,77 +299,78 @@ def display_dataset_analysis(
col_splitter="data_train_test",
split_values=["test", "train"],
names=["Prediction dataset", "Training dataset"],
- group_id='target'
+ group_id="target",
)
if multivariate_analysis:
print_md("### Multivariate analysis")
fig_corr = self.explainer.plot.correlations(
self.df_train_test,
- facet_col='data_train_test',
+ facet_col="data_train_test",
max_features=20,
- width=900 if len(self.df_train_test['data_train_test'].unique()) > 1 else 500,
+ width=900 if len(self.df_train_test["data_train_test"].unique()) > 1 else 500,
height=500,
)
print_html(plotly.io.to_html(fig_corr))
- print_md('---')
+ print_md("---")
def _display_dataset_analysis_global(self):
- df_stats_global = self._stats_to_table(test_stats=perform_global_dataframe_analysis(self.x_init),
- train_stats=perform_global_dataframe_analysis(self.x_train_pre),
- names=["Prediction dataset", "Training dataset"])
+ df_stats_global = self._stats_to_table(
+ test_stats=perform_global_dataframe_analysis(self.x_init),
+ train_stats=perform_global_dataframe_analysis(self.x_train_pre),
+ names=["Prediction dataset", "Training dataset"],
+ )
print_html(df_stats_global.to_html(classes="greyGridTable"))
def _perform_and_display_analysis_univariate(
- self,
- df: pd.DataFrame,
- col_splitter: str,
- split_values: list,
- names: list,
- group_id: str
+ self, df: pd.DataFrame, col_splitter: str, split_values: list, names: list, group_id: str
):
col_types = compute_col_types(df)
n_splits = df[col_splitter].nunique()
inv_columns_dict = {v: k for k, v in self.explainer.columns_dict.items()}
- test_stats_univariate = perform_univariate_dataframe_analysis(df.loc[df[col_splitter] == split_values[0]],
- col_types=col_types)
+ test_stats_univariate = perform_univariate_dataframe_analysis(
+ df.loc[df[col_splitter] == split_values[0]], col_types=col_types
+ )
if n_splits > 1:
- train_stats_univariate = perform_univariate_dataframe_analysis(df.loc[df[col_splitter] == split_values[1]],
- col_types=col_types)
+ train_stats_univariate = perform_univariate_dataframe_analysis(
+ df.loc[df[col_splitter] == split_values[1]], col_types=col_types
+ )
univariate_template = template_env.get_template("univariate.html")
univariate_features_desc = list()
- list_cols_labels = [self.explainer.features_dict.get(col, col)
- for col in df.drop(col_splitter, axis=1).columns.to_list()]
+ list_cols_labels = [
+ self.explainer.features_dict.get(col, col) for col in df.drop(col_splitter, axis=1).columns.to_list()
+ ]
for col_label in sorted(list_cols_labels):
col = self.explainer.inv_features_dict.get(col_label, col_label)
- fig = generate_fig_univariate(df_all=df, col=col, hue=col_splitter, type=col_types[col],
- colors_dict=self.explainer.colors_dict)
+ fig = generate_fig_univariate(
+ df_all=df, col=col, hue=col_splitter, type=col_types[col], colors_dict=self.explainer.colors_dict
+ )
df_col_stats = self._stats_to_table(
test_stats=test_stats_univariate[col],
train_stats=train_stats_univariate[col] if n_splits > 1 else None,
- names=names
+ names=names,
)
- univariate_features_desc.append({
- 'feature_index': int(inv_columns_dict.get(col, 0)),
- 'name': col,
- 'type': str(series_dtype(df[col])),
- 'description': col_label,
- 'table': df_col_stats.to_html(classes="greyGridTable"),
- 'image': convert_fig_to_html(fig)
- })
+ univariate_features_desc.append(
+ {
+ "feature_index": int(inv_columns_dict.get(col, 0)),
+ "name": col,
+ "type": str(series_dtype(df[col])),
+ "description": col_label,
+ "table": df_col_stats.to_html(classes="greyGridTable"),
+ "image": convert_fig_to_html(fig),
+ }
+ )
print_html(univariate_template.render(features=univariate_features_desc, groupId=group_id))
@staticmethod
- def _stats_to_table(test_stats: dict,
- names: list,
- train_stats: Optional[dict] = None,
- ) -> pd.DataFrame:
+ def _stats_to_table(
+ test_stats: dict,
+ names: list,
+ train_stats: Optional[dict] = None,
+ ) -> pd.DataFrame:
if train_stats is not None:
- return pd.DataFrame({
- names[1]: pd.Series(train_stats),
- names[0]: pd.Series(test_stats)
- })
+ return pd.DataFrame({names[1]: pd.Series(train_stats), names[0]: pd.Series(test_stats)})
else:
return pd.DataFrame({names[0]: pd.Series(test_stats)})
@@ -364,30 +385,34 @@ def display_model_explainability(self):
multiclass = True if (self.explainer._classes and len(self.explainer._classes) > 2) else False
c_list = self.explainer._classes if multiclass else [1] # list just used for multiclass
for index_label, label in enumerate(c_list): # Iterating over all labels in multiclass case
- label_value = self.explainer.check_label_name(label)[2] if multiclass else ''
+ label_value = self.explainer.check_label_name(label)[2] if multiclass else ""
fig_features_importance = self.explainer.plot.features_importance(label=label)
explain_contrib_data = list()
- list_cols_labels = [self.explainer.features_dict.get(col, col)
- for col in self.col_names]
+ list_cols_labels = [self.explainer.features_dict.get(col, col) for col in self.col_names]
for feature_label in sorted(list_cols_labels):
feature = self.explainer.inv_features_dict.get(feature_label, feature_label)
fig = self.explainer.plot.contribution_plot(feature, label=label, max_points=200)
- explain_contrib_data.append({
- 'feature_index': int(inv_columns_dict[feature]),
- 'name': feature,
- 'description': self.explainer.features_dict[feature],
- 'plot': plotly.io.to_html(fig, include_plotlyjs=False, full_html=False)
- })
- explain_data.append({
- 'index': index_label,
- 'name': label_value,
- 'feature_importance_plot': plotly.io.to_html(fig_features_importance, include_plotlyjs=False,
- full_html=False),
- 'features': explain_contrib_data
- })
+ explain_contrib_data.append(
+ {
+ "feature_index": int(inv_columns_dict[feature]),
+ "name": feature,
+ "description": self.explainer.features_dict[feature],
+ "plot": plotly.io.to_html(fig, include_plotlyjs=False, full_html=False),
+ }
+ )
+ explain_data.append(
+ {
+ "index": index_label,
+ "name": label_value,
+ "feature_importance_plot": plotly.io.to_html(
+ fig_features_importance, include_plotlyjs=False, full_html=False
+ ),
+ "features": explain_contrib_data,
+ }
+ )
print_html(explainability_template.render(labels=explain_data))
- print_md('---')
+ print_md("---")
def display_model_performance(self):
"""
@@ -418,39 +443,48 @@ def display_model_performance(self):
return
print_md("### Univariate analysis of target variable")
- df = pd.concat([pd.DataFrame({self.target_name: self.y_pred}).assign(_dataset="pred"),
- pd.DataFrame({self.target_name: self.y_test}).assign(_dataset="true")
- if self.y_test is not None else None]).reset_index(drop=True)
+ df = pd.concat(
+ [
+ pd.DataFrame({self.target_name: self.y_pred}).assign(_dataset="pred"),
+ pd.DataFrame({self.target_name: self.y_test}).assign(_dataset="true")
+ if self.y_test is not None
+ else None,
+ ]
+ ).reset_index(drop=True)
self._perform_and_display_analysis_univariate(
df=df,
col_splitter="_dataset",
split_values=["pred", "true"],
names=["Prediction values", "True values"],
- group_id='target-distribution'
+ group_id="target-distribution",
)
- if 'metrics' not in self.config.keys():
+ if "metrics" not in self.config.keys():
logging.info("No 'metrics' key found in report config dict. Skipping model performance part.")
return
print_md("### Metrics")
- for metric in self.config['metrics']:
- if 'name' not in metric.keys():
- metric['name'] = metric['path']
+ for metric in self.config["metrics"]:
+ if "name" not in metric.keys():
+ metric["name"] = metric["path"]
- if metric['path'] in ['confusion_matrix', 'sklearn.metrics.confusion_matrix'] or \
- metric['name'] == 'confusion_matrix':
+ if (
+ metric["path"] in ["confusion_matrix", "sklearn.metrics.confusion_matrix"]
+ or metric["name"] == "confusion_matrix"
+ ):
print_md(f"**{metric['name']} :**")
- print_html(convert_fig_to_html(generate_confusion_matrix_plot(
- y_true=self.y_test,
- y_pred=self.y_pred,
- colors_dict=self.explainer.colors_dict
- )))
+ print_html(
+ convert_fig_to_html(
+ generate_confusion_matrix_plot(
+ y_true=self.y_test, y_pred=self.y_pred, colors_dict=self.explainer.colors_dict
+ )
+ )
+ )
else:
try:
- metric_fn = get_callable(path=metric['path'])
+ metric_fn = get_callable(path=metric["path"])
# Look if we should use proba values instead of predicted values
- if 'use_proba_values' in metric.keys() and metric['use_proba_values'] is True:
+ if "use_proba_values" in metric.keys() and metric["use_proba_values"] is True:
y_pred = self.explainer.proba_values
else:
y_pred = self.y_pred
@@ -468,6 +502,8 @@ def display_model_performance(self):
print_md(f"**{metric['name']} :**")
print_html(f"{res} ")
else:
- logging.info(f"Could not compute following metric : {metric['path']}. \n"
- f"Result of type {res} cannot be displayed")
- print_md('---')
+ logging.info(
+ f"Could not compute following metric : {metric['path']}. \n"
+ f"Result of type {res} cannot be displayed"
+ )
+ print_md("---")
diff --git a/shapash/report/template/custom/conf.json b/shapash/report/template/custom/conf.json
index 76a9c7c5..a5974f9f 100644
--- a/shapash/report/template/custom/conf.json
+++ b/shapash/report/template/custom/conf.json
@@ -1,12 +1,12 @@
{
- "base_template": "classic",
- "mimetypes": {
- "text/html": true
- },
- "preprocessors": {
- "100-pygments": {
- "type": "nbconvert.preprocessors.CSSHTMLHeaderPreprocessor",
- "enabled": true
- }
+ "base_template": "classic",
+ "mimetypes": {
+ "text/html": true
+ },
+ "preprocessors": {
+ "100-pygments": {
+ "enabled": true,
+ "type": "nbconvert.preprocessors.CSSHTMLHeaderPreprocessor"
}
-}
\ No newline at end of file
+ }
+}
diff --git a/shapash/report/visualisation.py b/shapash/report/visualisation.py
index b82d0d8d..66a11b2a 100644
--- a/shapash/report/visualisation.py
+++ b/shapash/report/visualisation.py
@@ -1,7 +1,6 @@
-from IPython.display import display, Markdown, Latex, HTML
import matplotlib.pyplot as plt
-
import pandas as pd
+from IPython.display import HTML, Latex, Markdown, display
def print_md(text: str):
@@ -26,7 +25,8 @@ def print_html(text: str):
def print_css_style():
- print_html("""
+ print_html(
+ """
- """)
+ """
+ )
def print_javascript_misc():
- print_html("""
+ print_html(
+ """
- """)
+ """
+ )
def convert_fig_to_html(fig):
- """ Convert Matplotlib figure 'fig' into a "
category_encoder_ordinal = ""
@@ -12,18 +12,17 @@
category_encoder_binary = ""
category_encoder_targetencoder = ""
-dummies_category_encoder = (category_encoder_onehot,
- category_encoder_binary,
- category_encoder_basen)
+dummies_category_encoder = (category_encoder_onehot, category_encoder_binary, category_encoder_basen)
-no_dummies_category_encoder = (category_encoder_ordinal,
- category_encoder_targetencoder)
+no_dummies_category_encoder = (category_encoder_ordinal, category_encoder_targetencoder)
-supported_category_encoder = (category_encoder_onehot,
- category_encoder_binary,
- category_encoder_basen,
- category_encoder_ordinal,
- category_encoder_targetencoder)
+supported_category_encoder = (
+ category_encoder_onehot,
+ category_encoder_binary,
+ category_encoder_basen,
+ category_encoder_ordinal,
+ category_encoder_targetencoder,
+)
def inv_transform_ce(x_in, encoding):
@@ -55,7 +54,7 @@ def inv_transform_ce(x_in, encoding):
rst = inv_transform_ordinal(x, encoding.ordinal_encoder.mapping)
elif str(type(encoding)) == category_encoder_binary:
- if ce.__version__ <= '2.2.2':
+ if ce.__version__ <= "2.2.2":
x = reverse_basen(x_in, encoding.base_n_encoder)
rst = inv_transform_ordinal(x, encoding.base_n_encoder.ordinal_encoder.mapping)
else:
@@ -104,20 +103,22 @@ def inv_transform_target(x_in, enc_target):
The reversed dataframe.
"""
for tgt_enc in enc_target.ordinal_encoder.mapping:
- name_target = tgt_enc.get('col')
+ name_target = tgt_enc.get("col")
mapping_ordinal = enc_target.mapping[name_target]
- mapping_target = tgt_enc.get('mapping')
+ mapping_target = tgt_enc.get("mapping")
reverse_target = pd.Series(mapping_target.index.values, index=mapping_target)
- rst_target = pd.concat([reverse_target, mapping_ordinal], axis=1, join='inner').fillna(value='NaN')
- aggregate = rst_target.groupby(1)[0].apply(lambda x: ' / '.join(map(str, x)))
+ rst_target = pd.concat([reverse_target, mapping_ordinal], axis=1, join="inner").fillna(value="NaN")
+ aggregate = rst_target.groupby(1)[0].apply(lambda x: " / ".join(map(str, x)))
if aggregate.shape[0] != rst_target.shape[0]:
- raise Exception('Multiple label found for the same value in TargetEncoder on col '+str(name_target) + '.')
+ raise Exception("Multiple label found for the same value in TargetEncoder on col " + str(name_target) + ".")
# print("Warning in inverse TargetEncoder - col " + str(name_target) + ": Multiple label for the same value, "
# "each label will be separate using : / ")
- transco = {'col': name_target,
- 'mapping': pd.Series(data=aggregate.index, index=aggregate.values),
- 'data_type': 'object'}
+ transco = {
+ "col": name_target,
+ "mapping": pd.Series(data=aggregate.index, index=aggregate.values),
+ "data_type": "object",
+ }
x_in = inv_transform_ordinal(x_in, [transco])
return x_in
@@ -139,15 +140,15 @@ def inv_transform_ordinal(x_in, encoding):
The reversed dataframe.
"""
for switch in encoding:
- col_name = switch.get('col')
- if not col_name in x_in.columns:
- raise Exception(f'Columns {col_name} not in dataframe.')
- column_mapping = switch.get('mapping')
+ col_name = switch.get("col")
+ if col_name not in x_in.columns:
+ raise Exception(f"Columns {col_name} not in dataframe.")
+ column_mapping = switch.get("mapping")
if isinstance(column_mapping, dict):
inverse = pd.Series(data=column_mapping.keys(), index=column_mapping.values())
else:
inverse = pd.Series(data=column_mapping.index, index=column_mapping.values)
- x_in[col_name] = x_in[col_name].map(inverse).astype(switch.get('data_type'))
+ x_in[col_name] = x_in[col_name].map(inverse).astype(switch.get("data_type"))
return x_in
@@ -170,11 +171,11 @@ def reverse_basen(x_in, encoding):
x = x_in.copy(deep=True)
out_cols = x.columns.values.tolist()
for ind_enc in range(len(encoding.mapping)):
- col_list = encoding.mapping[ind_enc].get('mapping').columns.tolist()
+ col_list = encoding.mapping[ind_enc].get("mapping").columns.tolist()
insert_at = out_cols.index(col_list[0])
if encoding.base == 1:
- value_array = np.array([int(col0.split('_')[-1]) for col0 in col_list])
+ value_array = np.array([int(col0.split("_")[-1]) for col0 in col_list])
else:
len0 = len(col_list)
value_array = np.array([encoding.base ** (len0 - 1 - i) for i in range(len0)])
@@ -205,14 +206,14 @@ def calc_inv_contrib_ce(x_contrib, encoding, agg_columns):
The aggregate contributions depending on which processing is apply.
"""
if str(type(encoding)) in dummies_category_encoder:
- if str(type(encoding)) in category_encoder_binary and ce.__version__ <= '2.2.2':
+ if str(type(encoding)) in category_encoder_binary and ce.__version__ <= "2.2.2":
encoding = encoding.base_n_encoder
drop_col = []
for switch in encoding.mapping:
- col_in = switch.get('col')
- mod = switch.get('mapping').columns.tolist()
+ col_in = switch.get("col")
+ mod = switch.get("mapping").columns.tolist()
insert_at = x_contrib.columns.tolist().index(mod[0])
- if agg_columns == 'first':
+ if agg_columns == "first":
x_contrib.insert(insert_at, col_in, x_contrib[mod[0]])
else:
x_contrib.insert(insert_at, col_in, x_contrib[mod].sum(axis=1))
@@ -240,9 +241,13 @@ def transform_ce(x_in, encoding):
pandas.Dataframe
The dataset preprocessed with the given encoding.
"""
- encoder = [category_encoder_ordinal, category_encoder_onehot,
- category_encoder_basen, category_encoder_binary,
- category_encoder_targetencoder]
+ encoder = [
+ category_encoder_ordinal,
+ category_encoder_onehot,
+ category_encoder_basen,
+ category_encoder_binary,
+ category_encoder_targetencoder,
+ ]
if str(type(encoding)) in encoder:
rst = encoding.transform(x_in)
@@ -273,15 +278,15 @@ def transform_ordinal(x_in, encoding):
The dataframe preprocessed.
"""
for switch in encoding:
- col_name = switch.get('col')
- if not col_name in x_in.columns:
- raise Exception(f'Columns {col_name} not in dataframe.')
- column_mapping = switch.get('mapping')
+ col_name = switch.get("col")
+ if col_name not in x_in.columns:
+ raise Exception(f"Columns {col_name} not in dataframe.")
+ column_mapping = switch.get("mapping")
if isinstance(column_mapping, dict):
transform = pd.Series(data=column_mapping.values(), index=column_mapping.keys())
else:
transform = pd.Series(data=column_mapping.values, index=column_mapping.index)
- x_in[col_name] = x_in[col_name].map(transform).astype(switch.get('mapping').values.dtype)
+ x_in[col_name] = x_in[col_name].map(transform).astype(switch.get("mapping").values.dtype)
return x_in
@@ -299,11 +304,15 @@ def get_col_mapping_ce(encoder):
dict_col_mapping : dict
Dict of mapping between dataframe columns before and after encoding.
"""
- if str(type(encoder)) in [category_encoder_ordinal, category_encoder_onehot, category_encoder_basen,
- category_encoder_targetencoder]:
+ if str(type(encoder)) in [
+ category_encoder_ordinal,
+ category_encoder_onehot,
+ category_encoder_basen,
+ category_encoder_targetencoder,
+ ]:
encoder_mapping = encoder.mapping
elif str(type(encoder)) == category_encoder_binary:
- if ce.__version__ <= '2.2.2':
+ if ce.__version__ <= "2.2.2":
encoder_mapping = encoder.base_n_encoder.mapping
else:
encoder_mapping = encoder.mapping
@@ -316,10 +325,10 @@ def get_col_mapping_ce(encoder):
dict_col_mapping[col] = [col]
elif isinstance(encoder_mapping, list):
for col_enc in encoder_mapping:
- if isinstance(col_enc.get('mapping'), pd.DataFrame):
- dict_col_mapping[col_enc.get('col')] = col_enc.get('mapping').columns.to_list()
+ if isinstance(col_enc.get("mapping"), pd.DataFrame):
+ dict_col_mapping[col_enc.get("col")] = col_enc.get("mapping").columns.to_list()
else:
- dict_col_mapping[col_enc.get('col')] = [col_enc.get('col')]
+ dict_col_mapping[col_enc.get("col")] = [col_enc.get("col")]
else:
raise NotImplementedError
return dict_col_mapping
diff --git a/shapash/utils/check.py b/shapash/utils/check.py
index 2dcdb9ed..69892339 100644
--- a/shapash/utils/check.py
+++ b/shapash/utils/check.py
@@ -2,15 +2,25 @@
Check Module
"""
import copy
+
import numpy as np
import pandas as pd
-from shapash.utils.category_encoder_backend import no_dummies_category_encoder, supported_category_encoder,\
- dummies_category_encoder
-from shapash.utils.columntransformer_backend import no_dummies_sklearn, supported_sklearn
+
+from shapash.utils.category_encoder_backend import (
+ dummies_category_encoder,
+ no_dummies_category_encoder,
+ supported_category_encoder,
+)
+from shapash.utils.columntransformer_backend import (
+ columntransformer,
+ get_feature_names,
+ get_list_features_names,
+ no_dummies_sklearn,
+ supported_sklearn,
+)
from shapash.utils.model import extract_features_model
from shapash.utils.model_synoptic import dict_model_feature
-from shapash.utils.transform import preprocessing_tolist, check_transformers
-from shapash.utils.columntransformer_backend import columntransformer, get_feature_names, get_list_features_names
+from shapash.utils.transform import check_transformers, preprocessing_tolist
def check_preprocessing(preprocessing=None):
@@ -27,6 +37,7 @@ def check_preprocessing(preprocessing=None):
use_ct, use_ce = check_transformers(list_preprocessing)
return use_ct, use_ce
+
def check_model(model):
"""
Check if model has a predict_proba method is a one column dataframe of integer or float
@@ -43,25 +54,25 @@ def check_model(model):
'regression' or 'classification' according to the attributes of the model
"""
_classes = None
- if hasattr(model, 'predict'):
- if hasattr(model, 'predict_proba') or \
- any(hasattr(model, attrib) for attrib in ['classes_', '_classes']):
- if hasattr(model, '_classes'): _classes = model._classes
- if hasattr(model, 'classes_'): _classes = model.classes_
- if isinstance(_classes, np.ndarray): _classes = _classes.tolist()
- if hasattr(model, 'predict_proba') and _classes == []: _classes = [0, 1] # catboost binary
- if hasattr(model, 'predict_proba') and _classes is None:
- raise ValueError(
- "No attribute _classes, classification model not supported"
- )
+ if hasattr(model, "predict"):
+ if hasattr(model, "predict_proba") or any(hasattr(model, attrib) for attrib in ["classes_", "_classes"]):
+ if hasattr(model, "_classes"):
+ _classes = model._classes
+ if hasattr(model, "classes_"):
+ _classes = model.classes_
+ if isinstance(_classes, np.ndarray):
+ _classes = _classes.tolist()
+ if hasattr(model, "predict_proba") and _classes == []:
+ _classes = [0, 1] # catboost binary
+ if hasattr(model, "predict_proba") and _classes is None:
+ raise ValueError("No attribute _classes, classification model not supported")
if _classes not in (None, []):
- return 'classification', _classes
+ return "classification", _classes
else:
- return 'regression', None
+ return "regression", None
else:
- raise ValueError(
- "No method predict in the specified model. Please, check model parameter"
- )
+ raise ValueError("No method predict in the specified model. Please, check model parameter")
+
def check_label_dict(label_dict, case, classes=None):
"""
@@ -76,14 +87,15 @@ def check_label_dict(label_dict, case, classes=None):
classes: list, None
List of labels if the model used is for classification problem, None otherwise.
"""
- if label_dict is not None and case == 'classification':
+ if label_dict is not None and case == "classification":
if set(classes) != set(list(label_dict.keys())):
raise ValueError(
- "label_dict and don't match: \n" +
- f"label_dict keys: {str(list(label_dict.keys()))}\n" +
- f"Classes model values {str(classes)}"
+ "label_dict and don't match: \n"
+ + f"label_dict keys: {str(list(label_dict.keys()))}\n"
+ + f"Classes model values {str(classes)}"
)
+
def check_mask_params(mask_params):
"""
Check if mask_params given respect the expected format.
@@ -96,24 +108,24 @@ def check_mask_params(mask_params):
if not isinstance(mask_params, dict):
raise ValueError(
"""
- mask_params must be a dict
+ mask_params must be a dict
"""
)
else:
conform_arguments = ["features_to_hide", "threshold", "positive", "max_contrib"]
- mask_arguments_not_conform = [argument for argument in mask_params.keys()
- if argument not in conform_arguments]
+ mask_arguments_not_conform = [argument for argument in mask_params.keys() if argument not in conform_arguments]
if len(mask_arguments_not_conform) != 0:
raise ValueError(
- """
+ """
mask_params must only have the following key arguments:
-feature_to_hide
-threshold
-positive
- -max_contrib
+ -max_contrib
"""
)
+
def check_y(x=None, y=None, y_name="y_target"):
"""
Check that ypred given has the right shape and expected value.
@@ -135,16 +147,17 @@ def check_y(x=None, y=None, y_name="y_target"):
if isinstance(y, pd.DataFrame):
if y.shape[1] > 1:
raise ValueError(f"{y_name} must be a one column pd.Dataframe or pd.Series.")
- if (y.dtypes.iloc[0] not in [float, int, np.int32, np.float32, np.int64, np.float64]):
+ if y.dtypes.iloc[0] not in [float, int, np.int32, np.float32, np.int64, np.float64]:
raise ValueError(f"{y_name} must contain int or float only")
if isinstance(y, pd.Series):
- if (y.dtype not in [float, int, np.int32, np.float32, np.int64, np.float64]):
+ if y.dtype not in [float, int, np.int32, np.float32, np.int64, np.float64]:
raise ValueError(f"{y_name} must contain int or float only")
y = y.to_frame()
if isinstance(y.columns[0], (int, float)):
y.columns = [y_name]
return y
+
def check_contribution_object(case, classes, contributions):
"""
Check len of list if _case is "classification"
@@ -162,9 +175,9 @@ def check_contribution_object(case, classes, contributions):
if case == "regression" and isinstance(contributions, (np.ndarray, pd.DataFrame)) == False:
raise ValueError(
"""
- Type of contributions parameter specified is not compatible with
+ Type of contributions parameter specified is not compatible with
regression model.
- Please check model and contributions parameters.
+ Please check model and contributions parameters.
"""
)
elif case == "classification":
@@ -180,15 +193,24 @@ def check_contribution_object(case, classes, contributions):
else:
raise ValueError(
"""
- Type of contributions parameter specified is not compatible with
+ Type of contributions parameter specified is not compatible with
classification model.
Please check model and contributions parameters.
"""
)
-def check_consistency_model_features(features_dict, model, columns_dict, features_types,
- mask_params=None, preprocessing=None, postprocessing=None,
- list_preprocessing=None, features_groups=None):
+
+def check_consistency_model_features(
+ features_dict,
+ model,
+ columns_dict,
+ features_types,
+ mask_params=None,
+ preprocessing=None,
+ postprocessing=None,
+ list_preprocessing=None,
+ features_groups=None,
+):
"""
Check the matching between attributes, features names are same, or include
@@ -229,8 +251,8 @@ def check_consistency_model_features(features_dict, model, columns_dict, feature
raise ValueError("features of features_types and columns_dict must be the same")
if mask_params is not None:
- if mask_params['features_to_hide'] is not None:
- if not all(feature in set(features_types) for feature in mask_params['features_to_hide']):
+ if mask_params["features_to_hide"] is not None:
+ if not all(feature in set(features_types) for feature in mask_params["features_to_hide"]):
raise ValueError("All features of mask_params must be in model")
if preprocessing is not None and str(type(preprocessing)) in (supported_category_encoder):
@@ -252,22 +274,26 @@ def check_consistency_model_features(features_dict, model, columns_dict, feature
if set(columns_dict_feature) != set(feature_expected_model):
raise ValueError("Features of columns_dict and model must be the same.")
else:
- if len(set(columns_dict.values())) != model_expected :
+ if len(set(columns_dict.values())) != model_expected:
raise ValueError("Features of columns_dict and model must have the same length")
if str(type(preprocessing)) in supported_category_encoder and isinstance(feature_expected_model, list):
if set(preprocessing.feature_names_out_) != set(feature_expected_model):
- raise ValueError("""
+ raise ValueError(
+ """
One of features returned by the Category_Encoders preprocessing doesn't
match the model's expected features.
- """)
+ """
+ )
elif preprocessing is not None:
feature_encoded = get_list_features_names(list_preprocessing, columns_dict)
if model_expected != len(feature_encoded):
- raise ValueError("""
+ raise ValueError(
+ """
Number of features returned by the preprocessing step doesn't
match the model's expected features.
- """)
+ """
+ )
if postprocessing:
if not isinstance(postprocessing, dict):
@@ -279,6 +305,7 @@ def check_consistency_model_features(features_dict, model, columns_dict, feature
raise ValueError("Postprocessing and columns_dict must have the same features names.")
check_postprocessing(features_types, postprocessing)
+
def check_preprocessing_options(columns_dict, features_dict, preprocessing=None, list_preprocessing=None):
"""
Check if preprocessing for ColumnTransformer doesn't have "drop" option otherwise compute several
@@ -309,8 +336,7 @@ def check_preprocessing_options(columns_dict, features_dict, preprocessing=None,
if len(feature_to_drop) != 0:
feature_to_drop = [columns_dict[index] for index in feature_to_drop]
- features_dict_op = {key: value for key, value in features_dict.items()
- if key not in feature_to_drop}
+ features_dict_op = {key: value for key, value in features_dict.items() if key not in feature_to_drop}
i = 0
columns_dict_op = dict()
@@ -319,13 +345,16 @@ def check_preprocessing_options(columns_dict, features_dict, preprocessing=None,
columns_dict_op[i] = value
i += 1
- return {"features_to_drop": feature_to_drop,
- "features_dict_op": features_dict_op,
- "columns_dict_op": columns_dict_op}
+ return {
+ "features_to_drop": feature_to_drop,
+ "features_dict_op": features_dict_op,
+ "columns_dict_op": columns_dict_op,
+ }
else:
return None
+
def check_consistency_model_label(columns_dict, label_dict=None):
"""
Check the matching between attributes, features names are same, or include
@@ -342,6 +371,7 @@ def check_consistency_model_label(columns_dict, label_dict=None):
if not all(feat in columns_dict for feat in label_dict):
raise ValueError("All features of label_dict must be in model")
+
def check_postprocessing(x, postprocessing=None):
"""
Check that postprocessing parameter has good attributes matching with x dataset or with dict of types of
@@ -359,21 +389,22 @@ def check_postprocessing(x, postprocessing=None):
raise ValueError("Postprocessing parameter must be a dictionnary")
for key in postprocessing.keys():
-
dict_post = postprocessing[key]
if not isinstance(dict_post, dict):
raise ValueError(f"{key} values must be a dict")
- if not list(dict_post.keys()) == ['type', 'rule']:
+ if list(dict_post.keys()) != ["type", "rule"]:
raise ValueError("Wrong postprocessing keys, you need 'type' and 'rule' keys")
- if not dict_post['type'] in ['prefix', 'suffix', 'transcoding', 'regex', 'case']:
- raise ValueError("Wrong postprocessing method. \n"
- "The available methods are: 'prefix', 'suffix', 'transcoding', 'regex', or 'case'")
+ if dict_post["type"] not in ["prefix", "suffix", "transcoding", "regex", "case"]:
+ raise ValueError(
+ "Wrong postprocessing method. \n"
+ "The available methods are: 'prefix', 'suffix', 'transcoding', 'regex', or 'case'"
+ )
- if dict_post['type'] == 'case':
- if dict_post['rule'] not in ['lower', 'upper']:
+ if dict_post["type"] == "case":
+ if dict_post["rule"] not in ["lower", "upper"]:
raise ValueError("Case modification unknown. Available ones are 'lower', 'upper'.")
if isinstance(x, dict):
@@ -383,17 +414,20 @@ def check_postprocessing(x, postprocessing=None):
if not pd.api.types.is_string_dtype(x[key]):
raise ValueError(f"Expected string object to modify with upper/lower method in {key} dict")
- if dict_post['type'] == 'regex':
- if not set(dict_post['rule'].keys()) == {'in', 'out'}:
- raise ValueError(f"Regex modifications for {key} are not possible, the keys in 'rule' dict"
- f" must be 'in' and 'out'.")
- if isinstance(x,dict):
+ if dict_post["type"] == "regex":
+ if set(dict_post["rule"].keys()) != {"in", "out"}:
+ raise ValueError(
+ f"Regex modifications for {key} are not possible, the keys in 'rule' dict"
+ f" must be 'in' and 'out'."
+ )
+ if isinstance(x, dict):
if x[key] != "object":
raise ValueError(f"Expected string object to modify with regex methods in {key} dict")
else:
if not pd.api.types.is_string_dtype(x[key]):
raise ValueError(f"Expected string object to modify with upper/lower method in {key} dict")
+
def check_features_name(columns_dict, features_dict, features):
"""
Convert a list of feature names (string) or features ids into features ids.
@@ -426,9 +460,7 @@ def check_features_name(columns_dict, features_dict, features):
elif inv_columns_dict and all(f in columns_dict.values() for f in features):
features_ids = [inv_columns_dict[f] for f in features]
else:
- raise ValueError(
- 'All features must came from the same dict of features (technical names or domain names).'
- )
+ raise ValueError("All features must came from the same dict of features (technical names or domain names).")
else:
raise ValueError(
@@ -439,6 +471,7 @@ def check_features_name(columns_dict, features_dict, features):
)
return features_ids
+
def check_additional_data(x, additional_data):
if not isinstance(additional_data, pd.DataFrame):
raise ValueError(f"additional_data must be a pd.Dataframe.")
diff --git a/shapash/utils/columntransformer_backend.py b/shapash/utils/columntransformer_backend.py
index cdf83671..e788597b 100644
--- a/shapash/utils/columntransformer_backend.py
+++ b/shapash/utils/columntransformer_backend.py
@@ -2,17 +2,28 @@
sklearn columntransformer
"""
-import pandas as pd
import numpy as np
-from shapash.utils.category_encoder_backend import inv_transform_ordinal
-from shapash.utils.category_encoder_backend import inv_transform_ce
-from shapash.utils.category_encoder_backend import supported_category_encoder
-from shapash.utils.category_encoder_backend import dummies_category_encoder
-from shapash.utils.category_encoder_backend import category_encoder_binary
-from shapash.utils.category_encoder_backend import transform_ordinal, get_col_mapping_ce
-from shapash.utils.model_synoptic import simple_tree_model_sklearn, catboost_model,\
- linear_model, svm_model, xgboost_model, lightgbm_model, dict_model_feature
+import pandas as pd
+
+from shapash.utils.category_encoder_backend import (
+ category_encoder_binary,
+ dummies_category_encoder,
+ get_col_mapping_ce,
+ inv_transform_ce,
+ inv_transform_ordinal,
+ supported_category_encoder,
+ transform_ordinal,
+)
from shapash.utils.model import extract_features_model
+from shapash.utils.model_synoptic import (
+ catboost_model,
+ dict_model_feature,
+ lightgbm_model,
+ linear_model,
+ simple_tree_model_sklearn,
+ svm_model,
+ xgboost_model,
+)
columntransformer = ""
@@ -26,18 +37,17 @@
other_model = xgboost_model + catboost_model + lightgbm_model
-dummies_sklearn = (sklearn_onehot)
+dummies_sklearn = sklearn_onehot
-no_dummies_sklearn = (sklearn_ordinal,
- sklearn_standardscaler,
- sklearn_quantiletransformer,
- sklearn_powertransformer)
+no_dummies_sklearn = (sklearn_ordinal, sklearn_standardscaler, sklearn_quantiletransformer, sklearn_powertransformer)
-supported_sklearn = (sklearn_onehot,
- sklearn_ordinal,
- sklearn_standardscaler,
- sklearn_quantiletransformer,
- sklearn_powertransformer)
+supported_sklearn = (
+ sklearn_onehot,
+ sklearn_ordinal,
+ sklearn_standardscaler,
+ sklearn_quantiletransformer,
+ sklearn_powertransformer,
+)
def inv_transform_ct(x_in, encoding):
@@ -73,30 +83,22 @@ def inv_transform_ct(x_in, encoding):
col_encoding = enc[2]
# For Scikit encoding we use the associated inverse transform method
if str(type(ct_encoding)) in supported_sklearn:
- frame, init = inv_transform_sklearn_in_ct(x_in,
- init,
- name_encoding,
- col_encoding,
- ct_encoding)
+ frame, init = inv_transform_sklearn_in_ct(x_in, init, name_encoding, col_encoding, ct_encoding)
# For category encoding we use the mapping
elif str(type(ct_encoding)) in supported_category_encoder:
- frame, init = inv_transform_ce_in_ct(x_in,
- init,
- name_encoding,
- col_encoding,
- ct_encoding)
+ frame, init = inv_transform_ce_in_ct(x_in, init, name_encoding, col_encoding, ct_encoding)
# columns not encode
- elif name_encoding == 'remainder':
- if ct_encoding == 'passthrough':
+ elif name_encoding == "remainder":
+ if ct_encoding == "passthrough":
nb_col = len(col_encoding)
- frame = x_in.iloc[:, init:init + nb_col]
+ frame = x_in.iloc[:, init : init + nb_col]
else:
frame = pd.DataFrame()
else:
- raise Exception(f'{ct_encoding} is not supported yet.')
+ raise Exception(f"{ct_encoding} is not supported yet.")
rst = pd.concat([rst, frame], axis=1)
@@ -133,10 +135,10 @@ def inv_transform_ce_in_ct(x_in, init, name_encoding, col_encoding, ct_encoding)
init : np.int
Index of the last column use to make the transformation.
"""
- colname_output = [name_encoding + '_' + val for val in col_encoding]
+ colname_output = [name_encoding + "_" + val for val in col_encoding]
colname_input = ct_encoding.get_feature_names_out()
nb_col = len(colname_input)
- x_to_inverse = x_in.iloc[:, init:init + nb_col].copy()
+ x_to_inverse = x_in.iloc[:, init : init + nb_col].copy()
x_to_inverse.columns = colname_input
frame = inv_transform_ce(x_to_inverse, ct_encoding)
frame.columns = colname_output
@@ -168,13 +170,13 @@ def inv_transform_sklearn_in_ct(x_in, init, name_encoding, col_encoding, ct_enco
init : np.int
Index of the last column use to make the transformation.
"""
- colname_output = [name_encoding + '_' + val for val in col_encoding]
+ colname_output = [name_encoding + "_" + val for val in col_encoding]
if str(type(ct_encoding)) in dummies_sklearn:
colname_input = ct_encoding.get_feature_names_out(col_encoding)
nb_col = len(colname_input)
else:
nb_col = len(colname_output)
- x_inverse = ct_encoding.inverse_transform(x_in.iloc[:, init:init + nb_col])
+ x_inverse = ct_encoding.inverse_transform(x_in.iloc[:, init : init + nb_col])
frame = pd.DataFrame(x_inverse, columns=colname_output, index=x_in.index)
init += nb_col
return frame, init
@@ -212,9 +214,9 @@ def calc_inv_contrib_ct(x_contrib, encoding, agg_columns):
ct_encoding = enc[1]
col_encoding = enc[2]
- if str(type(ct_encoding)) in supported_category_encoder+supported_sklearn:
+ if str(type(ct_encoding)) in supported_category_encoder + supported_sklearn:
# We create new columns names depending on the name of the transformers and the name of the column.
- colname_output = [name_encoding + '_' + val for val in col_encoding]
+ colname_output = [name_encoding + "_" + val for val in col_encoding]
# If the processing create multiple columns we find the number of original categories and aggregate
# the contribution.
@@ -224,32 +226,32 @@ def calc_inv_contrib_ct(x_contrib, encoding, agg_columns):
col_origin = ct_encoding.categories_[i_enc]
elif str(type(ct_encoding)) == category_encoder_binary:
try:
- col_origin = ct_encoding.base_n_encoder.mapping[i_enc].get('mapping').columns.tolist()
+ col_origin = ct_encoding.base_n_encoder.mapping[i_enc].get("mapping").columns.tolist()
except:
- col_origin = ct_encoding.mapping[i_enc].get('mapping').columns.tolist()
+ col_origin = ct_encoding.mapping[i_enc].get("mapping").columns.tolist()
else:
- col_origin = ct_encoding.mapping[i_enc].get('mapping').columns.tolist()
+ col_origin = ct_encoding.mapping[i_enc].get("mapping").columns.tolist()
nb_col = len(col_origin)
- if agg_columns == 'first':
+ if agg_columns == "first":
contrib_inverse = x_contrib.iloc[:, init]
else:
- contrib_inverse = x_contrib.iloc[:, init:init + nb_col].sum(axis=1)
- frame = pd.DataFrame(contrib_inverse,
- columns=[colname_output[i_enc]],
- index=contrib_inverse.index)
+ contrib_inverse = x_contrib.iloc[:, init : init + nb_col].sum(axis=1)
+ frame = pd.DataFrame(
+ contrib_inverse, columns=[colname_output[i_enc]], index=contrib_inverse.index
+ )
rst = pd.concat([rst, frame], axis=1)
init += nb_col
else:
nb_col = len(colname_output)
- frame = x_contrib.iloc[:, init:init + nb_col]
+ frame = x_contrib.iloc[:, init : init + nb_col]
frame.columns = colname_output
rst = pd.concat([rst, frame], axis=1)
init += nb_col
- elif name_encoding == 'remainder':
- if ct_encoding == 'passthrough':
+ elif name_encoding == "remainder":
+ if ct_encoding == "passthrough":
nb_col = len(col_encoding)
- frame = x_contrib.iloc[:, init:init + nb_col]
+ frame = x_contrib.iloc[:, init : init + nb_col]
rst = pd.concat([rst, frame], axis=1)
init += nb_col
else:
@@ -286,14 +288,15 @@ def transform_ct(x_in, model, encoding):
if str(type(encoding)) == columntransformer:
# We use inverse tranform from the encoding method base on columns position
if str(type(model)) in sklearn_model:
- rst = pd.DataFrame(encoding.transform(x_in),
- index=x_in.index)
+ rst = pd.DataFrame(encoding.transform(x_in), index=x_in.index)
rst.columns = ["col_" + str(feature) for feature in rst.columns]
elif str(type(model)) in other_model:
- rst = pd.DataFrame(encoding.transform(x_in),
- columns=extract_features_model(model, dict_model_feature[str(type(model))]),
- index=x_in.index)
+ rst = pd.DataFrame(
+ encoding.transform(x_in),
+ columns=extract_features_model(model, dict_model_feature[str(type(model))]),
+ index=x_in.index,
+ )
else:
raise ValueError("Model specified isn't supported by Shapash.")
@@ -327,20 +330,18 @@ def get_names(name, trans, column, column_transformer):
list:
List of returned features when specific transformer is applied.
"""
- if trans == 'drop' or (
- hasattr(column, '__len__') and not len(column)):
+ if trans == "drop" or (hasattr(column, "__len__") and not len(column)):
return []
- if trans == 'passthrough':
- if hasattr(column_transformer, '_df_columns'):
- if ((not isinstance(column, slice))
- and all(isinstance(col, str) for col in column)):
+ if trans == "passthrough":
+ if hasattr(column_transformer, "_df_columns"):
+ if (not isinstance(column, slice)) and all(isinstance(col, str) for col in column):
return column
else:
return column_transformer._df_columns[column]
else:
indices = np.arange(column_transformer._n_features)
- return ['x%d' % i for i in indices[column]]
- if not hasattr(trans, 'get_feature_names_out'):
+ return ["x%d" % i for i in indices[column]]
+ if not hasattr(trans, "get_feature_names_out"):
if column is None:
return []
else:
@@ -408,9 +409,9 @@ def get_feature_out(estimator, feature_in):
"""
Returns estimator features out if it has get_feature_names_out method, else features_in
"""
- if hasattr(estimator, 'get_feature_names_out') and hasattr(estimator, 'categories_'):
+ if hasattr(estimator, "get_feature_names_out") and hasattr(estimator, "categories_"):
return estimator.get_feature_names_out(), estimator.categories_
- elif hasattr(estimator, 'get_feature_names_out'):
+ elif hasattr(estimator, "get_feature_names_out"):
return estimator.get_feature_names_out(), []
else:
return feature_in, []
@@ -435,38 +436,38 @@ def get_col_mapping_ct(encoder, x_encoded):
dict_col_mapping = dict()
idx_encoded = 0
for name, estimator, features in encoder.transformers_:
- if name != 'remainder':
+ if name != "remainder":
if str(type(estimator)) in dummies_sklearn:
features_out, categories_out = get_feature_out(estimator, features)
for i, f_name in enumerate(features):
- dict_col_mapping[name + '_' + f_name] = list()
+ dict_col_mapping[name + "_" + f_name] = list()
for _ in categories_out[i]:
- dict_col_mapping[name + '_' + f_name].append(x_encoded.columns.to_list()[idx_encoded])
+ dict_col_mapping[name + "_" + f_name].append(x_encoded.columns.to_list()[idx_encoded])
idx_encoded += 1
elif str(type(estimator)) in no_dummies_sklearn:
features_out, categories_out = get_feature_out(estimator, features)
for f_name in features_out:
- dict_col_mapping[name + '_' + f_name] = [x_encoded.columns.to_list()[idx_encoded]]
+ dict_col_mapping[name + "_" + f_name] = [x_encoded.columns.to_list()[idx_encoded]]
idx_encoded += 1
elif str(type(estimator)) in supported_category_encoder:
dict_mapping_ce = get_col_mapping_ce(estimator)
for f_name in dict_mapping_ce.keys():
- dict_col_mapping[name + '_' + f_name] = list()
+ dict_col_mapping[name + "_" + f_name] = list()
for _ in dict_mapping_ce[f_name]:
- dict_col_mapping[name + '_' + f_name].append(x_encoded.columns.to_list()[idx_encoded])
+ dict_col_mapping[name + "_" + f_name].append(x_encoded.columns.to_list()[idx_encoded])
idx_encoded += 1
else:
- raise NotImplementedError(f'Estimator not supported : {estimator}')
+ raise NotImplementedError(f"Estimator not supported : {estimator}")
- elif estimator == 'passthrough':
+ elif estimator == "passthrough":
try:
features_out = encoder.feature_names_in_[features]
except:
- features_out = encoder._feature_names_in[features] #for oldest sklearn version
+ features_out = encoder._feature_names_in[features] # for oldest sklearn version
for f_name in features_out:
dict_col_mapping[f_name] = [x_encoded.columns.to_list()[idx_encoded]]
idx_encoded += 1
diff --git a/shapash/utils/explanation_metrics.py b/shapash/utils/explanation_metrics.py
index 3e8d6bfe..04c18880 100644
--- a/shapash/utils/explanation_metrics.py
+++ b/shapash/utils/explanation_metrics.py
@@ -41,7 +41,7 @@ def _compute_distance(x1, x2, mean_vector, epsilon=0.0000001):
Returns
-------
diff : float
- Returns :math:`\\sum(\\frac{|x1-x2|}{mean\_vector+epsilon})`
+ Returns :math:`\\sum(\\frac{|x1-x2|}{mean\\_vector+epsilon})`
"""
diff = np.sum(np.abs(x1 - x2) / (mean_vector + epsilon))
return diff
@@ -109,7 +109,7 @@ def _get_radius(dataset, n_neighbors, sample_size=500, percentile=95):
similarity_distance[i, j] = dist
similarity_distance[j, i] = dist
# Select top n_neighbors
- ordered_X = np.sort(similarity_distance)[:, 1: n_neighbors + 1]
+ ordered_X = np.sort(similarity_distance)[:, 1 : n_neighbors + 1]
# Select the value of the distance that captures XX% of all distances (percentile)
return np.percentile(ordered_X.flatten(), percentile)
@@ -186,6 +186,7 @@ def find_neighbors(selection, dataset, model, mode, n_neighbors=10):
all_neighbors[i] = neighbors[neighbors[:, -2] < radius]
return all_neighbors
+
def shap_neighbors(instance, x_encoded, contributions, mode):
"""
For an instance and corresponding neighbors, calculate various
@@ -211,8 +212,11 @@ def shap_neighbors(instance, x_encoded, contributions, mode):
"""
# Extract SHAP values for instance and neighbors
# :-2 indicates that two columns are disregarded : distance to instance and model output
- ind = pd.merge(x_encoded.reset_index(), pd.DataFrame(instance[:, :-2], columns=x_encoded.columns), how='inner')\
- .set_index(x_encoded.index.name if x_encoded.index.name is not None else 'index').index
+ ind = (
+ pd.merge(x_encoded.reset_index(), pd.DataFrame(instance[:, :-2], columns=x_encoded.columns), how="inner")
+ .set_index(x_encoded.index.name if x_encoded.index.name is not None else "index")
+ .index
+ )
# If classification, select contrbutions of one class only
if mode == "classification" and len(contributions) == 2:
contributions = contributions[1]
@@ -223,12 +227,16 @@ def shap_neighbors(instance, x_encoded, contributions, mode):
norm_abs_shap_values = normalize(np.abs(shap_values), axis=1, norm="l1")
# Compute the average difference between the instance and its neighbors
# And replace NaN with 0
- average_diff = np.divide(norm_shap_values.std(axis=0), norm_abs_shap_values.mean(axis=0),
- out=np.zeros(norm_abs_shap_values.shape[1]),
- where=norm_abs_shap_values.mean(axis=0) != 0)
+ average_diff = np.divide(
+ norm_shap_values.std(axis=0),
+ norm_abs_shap_values.mean(axis=0),
+ out=np.zeros(norm_abs_shap_values.shape[1]),
+ where=norm_abs_shap_values.mean(axis=0) != 0,
+ )
return norm_shap_values, average_diff, norm_abs_shap_values[0, :]
+
def get_min_nb_features(selection, contributions, mode, distance):
"""
Determine the minimum number of features needed for the prediction \
@@ -237,13 +245,13 @@ def get_min_nb_features(selection, contributions, mode, distance):
The closeness is defined via the following distances:
- * For regression:
+ * For regression:
.. math::
-
+
distance = \\frac{|output_{allFeatures} - output_{currentFeatures}|}{|output_{allFeatures}|}
- * For classification:
+ * For classification:
.. math::
@@ -311,9 +319,9 @@ def get_distance(selection, contributions, mode, nb_features):
-------
distance : array
List of distances for each instance by using top selected features (ex: np.array([0.12, 0.16...])).
-
+
* For regression:
-
+
* normalized distance between the output of current model and output of full model
* For classifciation:
diff --git a/shapash/utils/io.py b/shapash/utils/io.py
index 51168c75..3f359290 100644
--- a/shapash/utils/io.py
+++ b/shapash/utils/io.py
@@ -2,8 +2,10 @@
IO module
"""
import pickle
+
try:
import yaml
+
_is_yaml_available = True
except (ImportError, ModuleNotFoundError):
_is_yaml_available = False
@@ -87,7 +89,7 @@ def load_yml(path):
"""
)
- with open(path, "r") as f:
+ with open(path) as f:
d = yaml.full_load(f)
return d
diff --git a/shapash/utils/load_smartpredictor.py b/shapash/utils/load_smartpredictor.py
index f0d5f99d..13946079 100644
--- a/shapash/utils/load_smartpredictor.py
+++ b/shapash/utils/load_smartpredictor.py
@@ -22,6 +22,4 @@ def load_smartpredictor(path):
if isinstance(predictor, SmartPredictor):
return predictor
else:
- raise ValueError(
- f"{predictor} is not an instance of type SmartPredictor"
- )
+ raise ValueError(f"{predictor} is not an instance of type SmartPredictor")
diff --git a/shapash/utils/model.py b/shapash/utils/model.py
index b3699caa..173837f8 100644
--- a/shapash/utils/model.py
+++ b/shapash/utils/model.py
@@ -2,8 +2,10 @@
Model Module
"""
from inspect import ismethod
+
import pandas as pd
+
def extract_features_model(model, model_attribute):
"""
Extract features of models if it's possible,
@@ -14,7 +16,7 @@ def extract_features_model(model, model_attribute):
model_attribute: String or List
if model can give features, attributes to access features, if not 'length'
"""
- if model_attribute[0] == 'length':
+ if model_attribute[0] == "length":
return model.n_features_in_
else:
if ismethod(getattr(model, model_attribute[0])):
@@ -45,11 +47,10 @@ def predict_proba(model, x_encoded, classes):
pandas.DataFrame
dataset of predicted proba for each label.
"""
- if hasattr(model, 'predict_proba'):
+ if hasattr(model, "predict_proba"):
proba_values = pd.DataFrame(
- model.predict_proba(x_encoded),
- columns=['class_' + str(x) for x in classes],
- index=x_encoded.index)
+ model.predict_proba(x_encoded), columns=["class_" + str(x) for x in classes], index=x_encoded.index
+ )
else:
raise ValueError("model has no predict_proba method")
@@ -72,8 +73,8 @@ def predict(model, x_encoded):
pandas.DataFrame
1-column dataframe containing the predictions.
"""
- if hasattr(model, 'predict'):
- y_pred = pd.DataFrame(model.predict(x_encoded), columns=['pred'], index=x_encoded.index)
+ if hasattr(model, "predict"):
+ y_pred = pd.DataFrame(model.predict(x_encoded), columns=["pred"], index=x_encoded.index)
else:
raise ValueError("model has no predict method")
@@ -100,10 +101,10 @@ def predict_error(y_target, y_pred, case):
1-column dataframe containing the prediction errors.
"""
prediction_error = None
- if y_target is not None and y_pred is not None and case=="regression":
+ if y_target is not None and y_pred is not None and case == "regression":
if (y_target == 0).any()[0]:
- prediction_error = abs(y_target.values-y_pred.values)
+ prediction_error = abs(y_target.values - y_pred.values)
else:
- prediction_error = abs((y_target.values-y_pred.values)/y_target.values)
+ prediction_error = abs((y_target.values - y_pred.values) / y_target.values)
prediction_error = pd.DataFrame(prediction_error, index=y_target.index, columns=["_error_"])
return prediction_error
diff --git a/shapash/utils/model_synoptic.py b/shapash/utils/model_synoptic.py
index ff8a24e1..a3f44d2f 100644
--- a/shapash/utils/model_synoptic.py
+++ b/shapash/utils/model_synoptic.py
@@ -3,55 +3,53 @@
"""
simple_tree_model_sklearn = (
- "",
- "",
- "",
- "",
- "",
- ""
- )
+ "",
+ "",
+ "",
+ "",
+ "",
+ "",
+)
xgboost_model = (
"",
"",
- "")
+ "",
+)
lightgbm_model = (
"",
"",
- ""
+ "",
)
-catboost_model = (
- "",
- "")
+catboost_model = ("", "")
linear_model = (
"",
- "")
+ "",
+)
-svm_model = (
- "",
- "")
+svm_model = ("", "")
simple_tree_model = simple_tree_model_sklearn + xgboost_model + lightgbm_model
-dict_model_feature = {"": ['length'],
- "": ['length'],
- "": ['length'],
- "": ['length'],
- "": ['length'],
- "": ['length'],
- "": ['length'],
- "": ['length'],
- "": ['length'],
- "": ['length'],
- "": ["booster_","feature_name"],
- "": ["booster_","feature_name"],
- "": ["feature_names"],
- "": ["get_booster","feature_names"],
- "": ["get_booster","feature_names"],
- "": ["feature_names"],
- "": ["feature_names_"],
- "": ["feature_names_"],
- }
-
+dict_model_feature = {
+ "": ["length"],
+ "": ["length"],
+ "": ["length"],
+ "": ["length"],
+ "": ["length"],
+ "": ["length"],
+ "": ["length"],
+ "": ["length"],
+ "": ["length"],
+ "": ["length"],
+ "": ["booster_", "feature_name"],
+ "": ["booster_", "feature_name"],
+ "": ["feature_names"],
+ "": ["get_booster", "feature_names"],
+ "": ["get_booster", "feature_names"],
+ "": ["feature_names"],
+ "": ["feature_names_"],
+ "": ["feature_names_"],
+}
diff --git a/shapash/utils/threading.py b/shapash/utils/threading.py
index d1c836d1..33065587 100644
--- a/shapash/utils/threading.py
+++ b/shapash/utils/threading.py
@@ -4,6 +4,7 @@
import sys
import threading
+
class CustomThread(threading.Thread):
"""
Python ovveride threading class
@@ -13,6 +14,7 @@ class CustomThread(threading.Thread):
threading : threading.Thread
Thread which you want to instanciate
"""
+
def __init__(self, *args, **keywords):
threading.Thread.__init__(self, *args, **keywords)
self.killed = False
@@ -23,7 +25,6 @@ def start(self):
self.run = self.__run
threading.Thread.start(self)
-
def __run(self):
sys.settrace(self.globaltrace)
self.__run_backup()
@@ -33,7 +34,7 @@ def globaltrace(self, frame, event, arg):
"""
Track the global trace
"""
- if event == 'call':
+ if event == "call":
return self.localtrace
else:
return None
@@ -43,7 +44,7 @@ def localtrace(self, frame, event, arg):
Track the local trace
"""
if self.killed:
- if event == 'line':
+ if event == "line":
raise SystemExit()
return self.localtrace
diff --git a/shapash/utils/transform.py b/shapash/utils/transform.py
index f55dfc22..877a044f 100644
--- a/shapash/utils/transform.py
+++ b/shapash/utils/transform.py
@@ -1,27 +1,30 @@
"""
Transform Module
"""
+import re
+
+import numpy as np
+import pandas as pd
+
+from shapash.utils.category_encoder_backend import (
+ get_col_mapping_ce,
+ inv_transform_ce,
+ supported_category_encoder,
+ transform_ce,
+)
from shapash.utils.columntransformer_backend import (
columntransformer,
+ get_col_mapping_ct,
inv_transform_ct,
supported_sklearn,
transform_ct,
- get_col_mapping_ct
)
-from shapash.utils.category_encoder_backend import (
- transform_ce,
- inv_transform_ce,
- supported_category_encoder,
- get_col_mapping_ce
-)
-import re
-import numpy as np
-import pandas as pd
# TODO
# encode targeted variable ? from sklearn.preprocessing import LabelEncoder
# make an easy version for dict, not writing all mapping
+
def inverse_transform(x_init, preprocessing=None):
"""
Reverse transformation giving a preprocessing.
@@ -72,6 +75,7 @@ def inverse_transform(x_init, preprocessing=None):
x_inverse = inv_transform_ce(x_inverse, encoding)
return x_inverse
+
def apply_preprocessing(x_init, model, preprocessing=None):
"""
Apply preprocessing on a raw dataset giving a preprocessing.
@@ -120,6 +124,7 @@ def apply_preprocessing(x_init, model, preprocessing=None):
x_init = transform_ce(x_init, encoding)
return x_init
+
def preprocessing_tolist(preprocess):
"""
Transform preprocess into a list, if preprocess contains a dict, transform the dict into a list of dict.
@@ -138,6 +143,7 @@ def preprocessing_tolist(preprocess):
list_encoding = [[x] if isinstance(x, dict) else x for x in list_encoding]
return list_encoding
+
def check_transformers(list_encoding):
"""
Check that all transformation are supported.
@@ -176,9 +182,10 @@ def check_transformers(list_encoding):
use_ct = True
for encoding in enc.transformers_:
ct_encoding = encoding[1]
- if (str(type(ct_encoding)) not in supported_sklearn) \
- and (str(type(ct_encoding)) not in supported_category_encoder):
- if str(type(ct_encoding)) != "" :
+ if (str(type(ct_encoding)) not in supported_sklearn) and (
+ str(type(ct_encoding)) not in supported_category_encoder
+ ):
+ if str(type(ct_encoding)) != "":
raise ValueError("One of the encoders used in ColumnTransformers isn't supported.")
elif str(type(enc)) in supported_category_encoder:
@@ -189,32 +196,32 @@ def check_transformers(list_encoding):
for enc_dict in enc:
if isinstance(enc_dict, dict):
# Check dict structure : col - mapping - data_type
- if not all(struct in enc_dict for struct in ('col', 'mapping', 'data_type')):
- raise Exception(f'{enc_dict} should have col, mapping and data_type as keys.')
+ if not all(struct in enc_dict for struct in ("col", "mapping", "data_type")):
+ raise Exception(f"{enc_dict} should have col, mapping and data_type as keys.")
else:
- raise Exception(f'{enc} is not a list of dict.')
+ raise Exception(f"{enc} is not a list of dict.")
else:
- raise Exception(f'{enc} is not supported yet.')
+ raise Exception(f"{enc} is not supported yet.")
# check that encoding don't use ColumnTransformer and Category encoding at the same time
if use_ct and use_ce:
raise Exception(
f"Can't support ColumnTransformer and Category encoding at the same time. "
- f"Use Category encoding in ColumnTransformer")
+ f"Use Category encoding in ColumnTransformer"
+ )
# check that Category encoding is apply on different columns
col = []
for enc in list_encoding:
- if not str(type(enc)) in ("",
- "",
- columntransformer):
+ if not str(type(enc)) in ("", "", columntransformer):
col += enc.cols
- duplicate = set([x for x in col if col.count(x) > 1])
+ duplicate = {x for x in col if col.count(x) > 1}
if duplicate:
- raise Exception('Columns ' + str(duplicate) + ' is used in multiple category encoding')
+ raise Exception("Columns " + str(duplicate) + " is used in multiple category encoding")
return use_ct, use_ce
+
def apply_postprocessing(x_init, postprocessing):
"""
Transforms x_init depending on postprocessing parameters.
@@ -237,36 +244,38 @@ def apply_postprocessing(x_init, postprocessing):
data_modif = new_preds[feature_name]
new_datai = list()
- if dict_postprocessing['type'] == 'prefix':
+ if dict_postprocessing["type"] == "prefix":
for value in data_modif.values:
- new_datai.append(dict_postprocessing['rule'] + str(value))
+ new_datai.append(dict_postprocessing["rule"] + str(value))
new_preds[feature_name] = new_datai
- elif dict_postprocessing['type'] == 'suffix':
+ elif dict_postprocessing["type"] == "suffix":
for value in data_modif.values:
- new_datai.append(str(value) + dict_postprocessing['rule'])
+ new_datai.append(str(value) + dict_postprocessing["rule"])
new_preds[feature_name] = new_datai
- elif dict_postprocessing['type'] == 'transcoding':
+ elif dict_postprocessing["type"] == "transcoding":
unique_values = x_init[feature_name].unique().tolist()
- unique_values = [value for value in unique_values if value not in dict_postprocessing['rule'].keys()]
+ unique_values = [value for value in unique_values if value not in dict_postprocessing["rule"].keys()]
for value in unique_values:
- dict_postprocessing['rule'][value] = value
- new_preds[feature_name] = new_preds[feature_name].map(dict_postprocessing['rule'])
+ dict_postprocessing["rule"][value] = value
+ new_preds[feature_name] = new_preds[feature_name].map(dict_postprocessing["rule"])
- elif dict_postprocessing['type'] == 'regex':
+ elif dict_postprocessing["type"] == "regex":
new_preds[feature_name] = new_preds[feature_name].apply(
- lambda x: re.sub(dict_postprocessing["rule"]['in'], dict_postprocessing["rule"]['out'], x))
+ lambda x: re.sub(dict_postprocessing["rule"]["in"], dict_postprocessing["rule"]["out"], x)
+ )
- elif dict_postprocessing['type'] == 'case':
- if dict_postprocessing['rule'] == 'lower':
+ elif dict_postprocessing["type"] == "case":
+ if dict_postprocessing["rule"] == "lower":
new_preds[feature_name] = new_preds[feature_name].apply(lambda x: x.lower())
- elif dict_postprocessing['rule'] == 'upper':
+ elif dict_postprocessing["rule"] == "upper":
new_preds[feature_name] = new_preds[feature_name].apply(lambda x: x.upper())
return new_preds
-def adapt_contributions(case,contributions):
+
+def adapt_contributions(case, contributions):
"""
If _case is "classification" and contributions a np.array or pd.DataFrame
this function transform contributions matrix in a list of 2 contributions
@@ -288,12 +297,8 @@ def adapt_contributions(case,contributions):
# So that we have the following format : [contributions_class_0, contributions_class_1, ...]
if isinstance(contributions, np.ndarray) and contributions.ndim == 3:
contributions = [contributions[:, :, i] for i in range(contributions.shape[-1])]
- if (
- (isinstance(contributions, pd.DataFrame) and case == 'classification')
- or (
- isinstance(contributions, (np.ndarray, list))
- and case == 'classification'
- and np.array(contributions).ndim == 2)
+ if (isinstance(contributions, pd.DataFrame) and case == "classification") or (
+ isinstance(contributions, (np.ndarray, list)) and case == "classification" and np.array(contributions).ndim == 2
):
return [contributions * -1, contributions]
else:
@@ -374,7 +379,8 @@ def get_features_transform_mapping(x_init, x_encoded, preprocessing=None):
dict_all_cols_mapping[col_name] = [col_name]
return dict_all_cols_mapping
-def handle_categorical_missing(df : pd.DataFrame)-> pd.DataFrame:
+
+def handle_categorical_missing(df: pd.DataFrame) -> pd.DataFrame:
"""
Replace missing values for categorical columns
@@ -383,7 +389,7 @@ def handle_categorical_missing(df : pd.DataFrame)-> pd.DataFrame:
df : pd.DataFrame
Pandas dataframe on which we will replace the missing values
"""
- categorical_cols = df.select_dtypes(include=['object']).columns
+ categorical_cols = df.select_dtypes(include=["object"]).columns
df_handle_missing = df.copy()
df_handle_missing[categorical_cols] = df_handle_missing[categorical_cols].fillna("missing")
return df_handle_missing
diff --git a/shapash/utils/translate.py b/shapash/utils/translate.py
index 776eac57..f5afeaf3 100644
--- a/shapash/utils/translate.py
+++ b/shapash/utils/translate.py
@@ -1,6 +1,8 @@
"""
Translate Module
"""
+
+
def translate(elements, mapping):
"""
Map a dictionary to a list of elements.
diff --git a/shapash/utils/utils.py b/shapash/utils/utils.py
index 6df58a4b..0fc4b673 100644
--- a/shapash/utils/utils.py
+++ b/shapash/utils/utils.py
@@ -1,12 +1,15 @@
"""
Utils is a group of function for the library
"""
+import math
+import socket
+
import numpy as np
import pandas as pd
-import socket
-import math
-from shapash.explainer.smart_state import SmartState
+
from shapash.explainer.multi_decorator import MultiDecorator
+from shapash.explainer.smart_state import SmartState
+
def get_host_name():
"""
@@ -68,6 +71,7 @@ def is_nested_list(object_param):
"""
return any(isinstance(elem, list) for elem in object_param)
+
def add_line_break(text, nbchar, maxlen=150):
"""
adding line break in string if necessary
@@ -86,7 +90,7 @@ def add_line_break(text, nbchar, maxlen=150):
string
original text + line break
"""
- if isinstance(text,str):
+ if isinstance(text, str):
length = 0
tot_length = 0
input_word = text.split()
@@ -97,20 +101,21 @@ def add_line_break(text, nbchar, maxlen=150):
if tot_length <= maxlen:
if length >= nbchar:
length = 0
- final_sep.append('+~]|"+M+")"+M+"*"),U=new RegExp(M+"|>"),X=new RegExp($),V=new RegExp("^"+I+"$"),G={ID:new RegExp("^#("+I+")"),CLASS:new RegExp("^\\.("+I+")"),TAG:new RegExp("^("+I+"|[*])"),ATTR:new RegExp("^"+W),PSEUDO:new RegExp("^"+$),CHILD:new RegExp("^:(only|first|last|nth|nth-last)-(child|of-type)(?:\\("+M+"*(even|odd|(([+-]|)(\\d*)n|)"+M+"*(?:([+-]|)"+M+"*(\\d+)|))"+M+"*\\)|)","i"),bool:new RegExp("^(?:"+R+")$","i"),needsContext:new RegExp("^"+M+"*[>+~]|:(even|odd|eq|gt|lt|nth|first|last)(?:\\("+M+"*((?:-\\d)?\\d*)"+M+"*\\)|)(?=[^-]|$)","i")},Y=/HTML$/i,Q=/^(?:input|select|textarea|button)$/i,J=/^h\d$/i,K=/^[^{]+\{\s*\[native \w/,Z=/^(?:#([\w-]+)|(\w+)|\.([\w-]+))$/,ee=/[+~]/,te=new RegExp("\\\\([\\da-f]{1,6}"+M+"?|("+M+")|.)","ig"),ne=function(e,t,n){var r="0x"+t-65536;return r!=r||n?t:r<0?String.fromCharCode(r+65536):String.fromCharCode(r>>10|55296,1023&r|56320)},re=/([\0-\x1f\x7f]|^-?\d)|^-$|[^\0-\x1f\x7f-\uFFFF\w-]/g,ie=function(e,t){return t?"\0"===e?"\ufffd":e.slice(0,-1)+"\\"+e.charCodeAt(e.length-1).toString(16)+" ":"\\"+e},oe=function(){T()},ae=be(function(e){return!0===e.disabled&&"fieldset"===e.nodeName.toLowerCase()},{dir:"parentNode",next:"legend"});try{H.apply(t=O.call(m.childNodes),m.childNodes),t[m.childNodes.length].nodeType}catch(e){H={apply:t.length?function(e,t){L.apply(e,O.call(t))}:function(e,t){var n=e.length,r=0;while(e[n++]=t[r++]);e.length=n-1}}}function se(t,e,n,r){var i,o,a,s,u,l,c,f=e&&e.ownerDocument,p=e?e.nodeType:9;if(n=n||[],"string"!=typeof t||!t||1!==p&&9!==p&&11!==p)return n;if(!r&&((e?e.ownerDocument||e:m)!==C&&T(e),e=e||C,E)){if(11!==p&&(u=Z.exec(t)))if(i=u[1]){if(9===p){if(!(a=e.getElementById(i)))return n;if(a.id===i)return n.push(a),n}else if(f&&(a=f.getElementById(i))&&y(e,a)&&a.id===i)return n.push(a),n}else{if(u[2])return H.apply(n,e.getElementsByTagName(t)),n;if((i=u[3])&&d.getElementsByClassName&&e.getElementsByClassName)return H.apply(n,e.getElementsByClassName(i)),n}if(d.qsa&&!A[t+" "]&&(!v||!v.test(t))&&(1!==p||"object"!==e.nodeName.toLowerCase())){if(c=t,f=e,1===p&&U.test(t)){(s=e.getAttribute("id"))?s=s.replace(re,ie):e.setAttribute("id",s=k),o=(l=h(t)).length;while(o--)l[o]="#"+s+" "+xe(l[o]);c=l.join(","),f=ee.test(t)&&ye(e.parentNode)||e}try{return H.apply(n,f.querySelectorAll(c)),n}catch(e){A(t,!0)}finally{s===k&&e.removeAttribute("id")}}}return g(t.replace(B,"$1"),e,n,r)}function ue(){var r=[];return function e(t,n){return r.push(t+" ")>b.cacheLength&&delete e[r.shift()],e[t+" "]=n}}function le(e){return e[k]=!0,e}function ce(e){var t=C.createElement("fieldset");try{return!!e(t)}catch(e){return!1}finally{t.parentNode&&t.parentNode.removeChild(t),t=null}}function fe(e,t){var n=e.split("|"),r=n.length;while(r--)b.attrHandle[n[r]]=t}function pe(e,t){var n=t&&e,r=n&&1===e.nodeType&&1===t.nodeType&&e.sourceIndex-t.sourceIndex;if(r)return r;if(n)while(n=n.nextSibling)if(n===t)return-1;return e?1:-1}function de(t){return function(e){return"input"===e.nodeName.toLowerCase()&&e.type===t}}function he(n){return function(e){var t=e.nodeName.toLowerCase();return("input"===t||"button"===t)&&e.type===n}}function ge(t){return function(e){return"form"in e?e.parentNode&&!1===e.disabled?"label"in e?"label"in e.parentNode?e.parentNode.disabled===t:e.disabled===t:e.isDisabled===t||e.isDisabled!==!t&&ae(e)===t:e.disabled===t:"label"in e&&e.disabled===t}}function ve(a){return le(function(o){return o=+o,le(function(e,t){var n,r=a([],e.length,o),i=r.length;while(i--)e[n=r[i]]&&(e[n]=!(t[n]=e[n]))})})}function ye(e){return e&&"undefined"!=typeof e.getElementsByTagName&&e}for(e in d=se.support={},i=se.isXML=function(e){var t=e.namespaceURI,n=(e.ownerDocument||e).documentElement;return!Y.test(t||n&&n.nodeName||"HTML")},T=se.setDocument=function(e){var t,n,r=e?e.ownerDocument||e:m;return r!==C&&9===r.nodeType&&r.documentElement&&(a=(C=r).documentElement,E=!i(C),m!==C&&(n=C.defaultView)&&n.top!==n&&(n.addEventListener?n.addEventListener("unload",oe,!1):n.attachEvent&&n.attachEvent("onunload",oe)),d.attributes=ce(function(e){return e.className="i",!e.getAttribute("className")}),d.getElementsByTagName=ce(function(e){return e.appendChild(C.createComment("")),!e.getElementsByTagName("*").length}),d.getElementsByClassName=K.test(C.getElementsByClassName),d.getById=ce(function(e){return a.appendChild(e).id=k,!C.getElementsByName||!C.getElementsByName(k).length}),d.getById?(b.filter.ID=function(e){var t=e.replace(te,ne);return function(e){return e.getAttribute("id")===t}},b.find.ID=function(e,t){if("undefined"!=typeof t.getElementById&&E){var n=t.getElementById(e);return n?[n]:[]}}):(b.filter.ID=function(e){var n=e.replace(te,ne);return function(e){var t="undefined"!=typeof e.getAttributeNode&&e.getAttributeNode("id");return t&&t.value===n}},b.find.ID=function(e,t){if("undefined"!=typeof t.getElementById&&E){var n,r,i,o=t.getElementById(e);if(o){if((n=o.getAttributeNode("id"))&&n.value===e)return[o];i=t.getElementsByName(e),r=0;while(o=i[r++])if((n=o.getAttributeNode("id"))&&n.value===e)return[o]}return[]}}),b.find.TAG=d.getElementsByTagName?function(e,t){return"undefined"!=typeof t.getElementsByTagName?t.getElementsByTagName(e):d.qsa?t.querySelectorAll(e):void 0}:function(e,t){var n,r=[],i=0,o=t.getElementsByTagName(e);if("*"===e){while(n=o[i++])1===n.nodeType&&r.push(n);return r}return o},b.find.CLASS=d.getElementsByClassName&&function(e,t){if("undefined"!=typeof t.getElementsByClassName&&E)return t.getElementsByClassName(e)},s=[],v=[],(d.qsa=K.test(C.querySelectorAll))&&(ce(function(e){a.appendChild(e).innerHTML="":{dir:"parentNode",first:!0}," ":{dir:"parentNode"},"+":{dir:"previousSibling",first:!0},"~":{dir:"previousSibling"}},preFilter:{ATTR:function(e){return e[1]=e[1].replace(te,ne),e[3]=(e[3]||e[4]||e[5]||"").replace(te,ne),"~="===e[2]&&(e[3]=" "+e[3]+" "),e.slice(0,4)},CHILD:function(e){return e[1]=e[1].toLowerCase(),"nth"===e[1].slice(0,3)?(e[3]||se.error(e[0]),e[4]=+(e[4]?e[5]+(e[6]||1):2*("even"===e[3]||"odd"===e[3])),e[5]=+(e[7]+e[8]||"odd"===e[3])):e[3]&&se.error(e[0]),e},PSEUDO:function(e){var t,n=!e[6]&&e[2];return G.CHILD.test(e[0])?null:(e[3]?e[2]=e[4]||e[5]||"":n&&X.test(n)&&(t=h(n,!0))&&(t=n.indexOf(")",n.length-t)-n.length)&&(e[0]=e[0].slice(0,t),e[2]=n.slice(0,t)),e.slice(0,3))}},filter:{TAG:function(e){var t=e.replace(te,ne).toLowerCase();return"*"===e?function(){return!0}:function(e){return e.nodeName&&e.nodeName.toLowerCase()===t}},CLASS:function(e){var t=p[e+" "];return t||(t=new RegExp("(^|"+M+")"+e+"("+M+"|$)"))&&p(e,function(e){return t.test("string"==typeof e.className&&e.className||"undefined"!=typeof e.getAttribute&&e.getAttribute("class")||"")})},ATTR:function(n,r,i){return function(e){var t=se.attr(e,n);return null==t?"!="===r:!r||(t+="","="===r?t===i:"!="===r?t!==i:"^="===r?i&&0===t.indexOf(i):"*="===r?i&&-1:\x20\t\r\n\f]*)[\x20\t\r\n\f]*\/?>(?:<\/\1>|)$/i;function j(e,n,r){return m(n)?k.grep(e,function(e,t){return!!n.call(e,t,e)!==r}):n.nodeType?k.grep(e,function(e){return e===n!==r}):"string"!=typeof n?k.grep(e,function(e){return-1)[^>]*|#([\w-]+))$/;(k.fn.init=function(e,t,n){var r,i;if(!e)return this;if(n=n||q,"string"==typeof e){if(!(r="<"===e[0]&&">"===e[e.length-1]&&3<=e.length?[null,e,null]:L.exec(e))||!r[1]&&t)return!t||t.jquery?(t||n).find(e):this.constructor(t).find(e);if(r[1]){if(t=t instanceof k?t[0]:t,k.merge(this,k.parseHTML(r[1],t&&t.nodeType?t.ownerDocument||t:E,!0)),D.test(r[1])&&k.isPlainObject(t))for(r in t)m(this[r])?this[r](t[r]):this.attr(r,t[r]);return this}return(i=E.getElementById(r[2]))&&(this[0]=i,this.length=1),this}return e.nodeType?(this[0]=e,this.length=1,this):m(e)?void 0!==n.ready?n.ready(e):e(k):k.makeArray(e,this)}).prototype=k.fn,q=k(E);var H=/^(?:parents|prev(?:Until|All))/,O={children:!0,contents:!0,next:!0,prev:!0};function P(e,t){while((e=e[t])&&1!==e.nodeType);return e}k.fn.extend({has:function(e){var t=k(e,this),n=t.length;return this.filter(function(){for(var e=0;e\x20\t\r\n\f]*)/i,he=/^$|^module$|\/(?:java|ecma)script/i,ge={option:[1,""," "],thead:[1,""],col:[2,""],tr:[2,""],td:[3,""],_default:[0,"",""]};function ve(e,t){var n;return n="undefined"!=typeof e.getElementsByTagName?e.getElementsByTagName(t||"*"):"undefined"!=typeof e.querySelectorAll?e.querySelectorAll(t||"*"):[],void 0===t||t&&A(e,t)?k.merge([e],n):n}function ye(e,t){for(var n=0,r=e.length;nx",y.noCloneChecked=!!me.cloneNode(!0).lastChild.defaultValue;var Te=/^key/,Ce=/^(?:mouse|pointer|contextmenu|drag|drop)|click/,Ee=/^([^.]*)(?:\.(.+)|)/;function ke(){return!0}function Se(){return!1}function Ne(e,t){return e===function(){try{return E.activeElement}catch(e){}}()==("focus"===t)}function Ae(e,t,n,r,i,o){var a,s;if("object"==typeof t){for(s in"string"!=typeof n&&(r=r||n,n=void 0),t)Ae(e,s,n,r,t[s],o);return e}if(null==r&&null==i?(i=n,r=n=void 0):null==i&&("string"==typeof n?(i=r,r=void 0):(i=r,r=n,n=void 0)),!1===i)i=Se;else if(!i)return e;return 1===o&&(a=i,(i=function(e){return k().off(e),a.apply(this,arguments)}).guid=a.guid||(a.guid=k.guid++)),e.each(function(){k.event.add(this,t,i,r,n)})}function De(e,i,o){o?(Q.set(e,i,!1),k.event.add(e,i,{namespace:!1,handler:function(e){var t,n,r=Q.get(this,i);if(1&e.isTrigger&&this[i]){if(r.length)(k.event.special[i]||{}).delegateType&&e.stopPropagation();else if(r=s.call(arguments),Q.set(this,i,r),t=o(this,i),this[i](),r!==(n=Q.get(this,i))||t?Q.set(this,i,!1):n={},r!==n)return e.stopImmediatePropagation(),e.preventDefault(),n.value}else r.length&&(Q.set(this,i,{value:k.event.trigger(k.extend(r[0],k.Event.prototype),r.slice(1),this)}),e.stopImmediatePropagation())}})):void 0===Q.get(e,i)&&k.event.add(e,i,ke)}k.event={global:{},add:function(t,e,n,r,i){var o,a,s,u,l,c,f,p,d,h,g,v=Q.get(t);if(v){n.handler&&(n=(o=n).handler,i=o.selector),i&&k.find.matchesSelector(ie,i),n.guid||(n.guid=k.guid++),(u=v.events)||(u=v.events={}),(a=v.handle)||(a=v.handle=function(e){return"undefined"!=typeof k&&k.event.triggered!==e.type?k.event.dispatch.apply(t,arguments):void 0}),l=(e=(e||"").match(R)||[""]).length;while(l--)d=g=(s=Ee.exec(e[l])||[])[1],h=(s[2]||"").split(".").sort(),d&&(f=k.event.special[d]||{},d=(i?f.delegateType:f.bindType)||d,f=k.event.special[d]||{},c=k.extend({type:d,origType:g,data:r,handler:n,guid:n.guid,selector:i,needsContext:i&&k.expr.match.needsContext.test(i),namespace:h.join(".")},o),(p=u[d])||((p=u[d]=[]).delegateCount=0,f.setup&&!1!==f.setup.call(t,r,h,a)||t.addEventListener&&t.addEventListener(d,a)),f.add&&(f.add.call(t,c),c.handler.guid||(c.handler.guid=n.guid)),i?p.splice(p.delegateCount++,0,c):p.push(c),k.event.global[d]=!0)}},remove:function(e,t,n,r,i){var o,a,s,u,l,c,f,p,d,h,g,v=Q.hasData(e)&&Q.get(e);if(v&&(u=v.events)){l=(t=(t||"").match(R)||[""]).length;while(l--)if(d=g=(s=Ee.exec(t[l])||[])[1],h=(s[2]||"").split(".").sort(),d){f=k.event.special[d]||{},p=u[d=(r?f.delegateType:f.bindType)||d]||[],s=s[2]&&new RegExp("(^|\\.)"+h.join("\\.(?:.*\\.|)")+"(\\.|$)"),a=o=p.length;while(o--)c=p[o],!i&&g!==c.origType||n&&n.guid!==c.guid||s&&!s.test(c.namespace)||r&&r!==c.selector&&("**"!==r||!c.selector)||(p.splice(o,1),c.selector&&p.delegateCount--,f.remove&&f.remove.call(e,c));a&&!p.length&&(f.teardown&&!1!==f.teardown.call(e,h,v.handle)||k.removeEvent(e,d,v.handle),delete u[d])}else for(d in u)k.event.remove(e,d+t[l],n,r,!0);k.isEmptyObject(u)&&Q.remove(e,"handle events")}},dispatch:function(e){var t,n,r,i,o,a,s=k.event.fix(e),u=new Array(arguments.length),l=(Q.get(this,"events")||{})[s.type]||[],c=k.event.special[s.type]||{};for(u[0]=s,t=1;t\x20\t\r\n\f]*)[^>]*)\/>/gi,qe=/\s*$/g;function Oe(e,t){return A(e,"table")&&A(11!==t.nodeType?t:t.firstChild,"tr")&&k(e).children("tbody")[0]||e}function Pe(e){return e.type=(null!==e.getAttribute("type"))+"/"+e.type,e}function Re(e){return"true/"===(e.type||"").slice(0,5)?e.type=e.type.slice(5):e.removeAttribute("type"),e}function Me(e,t){var n,r,i,o,a,s,u,l;if(1===t.nodeType){if(Q.hasData(e)&&(o=Q.access(e),a=Q.set(t,o),l=o.events))for(i in delete a.handle,a.events={},l)for(n=0,r=l[i].length;n")},clone:function(e,t,n){var r,i,o,a,s,u,l,c=e.cloneNode(!0),f=oe(e);if(!(y.noCloneChecked||1!==e.nodeType&&11!==e.nodeType||k.isXMLDoc(e)))for(a=ve(c),r=0,i=(o=ve(e)).length;r").attr(n.scriptAttrs||{}).prop({charset:n.scriptCharset,src:n.url}).on("load error",i=function(e){r.remove(),i=null,e&&t("error"===e.type?404:200,e.type)}),E.head.appendChild(r[0])},abort:function(){i&&i()}}});var Vt,Gt=[],Yt=/(=)\?(?=&|$)|\?\?/;k.ajaxSetup({jsonp:"callback",jsonpCallback:function(){var e=Gt.pop()||k.expando+"_"+kt++;return this[e]=!0,e}}),k.ajaxPrefilter("json jsonp",function(e,t,n){var r,i,o,a=!1!==e.jsonp&&(Yt.test(e.url)?"url":"string"==typeof e.data&&0===(e.contentType||"").indexOf("application/x-www-form-urlencoded")&&Yt.test(e.data)&&"data");if(a||"jsonp"===e.dataTypes[0])return r=e.jsonpCallback=m(e.jsonpCallback)?e.jsonpCallback():e.jsonpCallback,a?e[a]=e[a].replace(Yt,"$1"+r):!1!==e.jsonp&&(e.url+=(St.test(e.url)?"&":"?")+e.jsonp+"="+r),e.converters["script json"]=function(){return o||k.error(r+" was not called"),o[0]},e.dataTypes[0]="json",i=C[r],C[r]=function(){o=arguments},n.always(function(){void 0===i?k(C).removeProp(r):C[r]=i,e[r]&&(e.jsonpCallback=t.jsonpCallback,Gt.push(r)),o&&m(i)&&i(o[0]),o=i=void 0}),"script"}),y.createHTMLDocument=((Vt=E.implementation.createHTMLDocument("").body).innerHTML="",2===Vt.childNodes.length),k.parseHTML=function(e,t,n){return"string"!=typeof e?[]:("boolean"==typeof t&&(n=t,t=!1),t||(y.createHTMLDocument?((r=(t=E.implementation.createHTMLDocument("")).createElement("base")).href=E.location.href,t.head.appendChild(r)):t=E),o=!n&&[],(i=D.exec(e))?[t.createElement(i[1])]:(i=we([e],t,o),o&&o.length&&k(o).remove(),k.merge([],i.childNodes)));var r,i,o},k.fn.load=function(e,t,n){var r,i,o,a=this,s=e.indexOf(" ");return-1").append(k.parseHTML(e)).find(r):e)}).always(n&&function(e,t){a.each(function(){n.apply(this,o||[e.responseText,t,e])})}),this},k.each(["ajaxStart","ajaxStop","ajaxComplete","ajaxError","ajaxSuccess","ajaxSend"],function(e,t){k.fn[t]=function(e){return this.on(t,e)}}),k.expr.pseudos.animated=function(t){return k.grep(k.timers,function(e){return t===e.elem}).length},k.offset={setOffset:function(e,t,n){var r,i,o,a,s,u,l=k.css(e,"position"),c=k(e),f={};"static"===l&&(e.style.position="relative"),s=c.offset(),o=k.css(e,"top"),u=k.css(e,"left"),("absolute"===l||"fixed"===l)&&-1<(o+u).indexOf("auto")?(a=(r=c.position()).top,i=r.left):(a=parseFloat(o)||0,i=parseFloat(u)||0),m(t)&&(t=t.call(e,n,k.extend({},s))),null!=t.top&&(f.top=t.top-s.top+a),null!=t.left&&(f.left=t.left-s.left+i),"using"in t?t.using.call(e,f):c.css(f)}},k.fn.extend({offset:function(t){if(arguments.length)return void 0===t?this:this.each(function(e){k.offset.setOffset(this,t,e)});var e,n,r=this[0];return r?r.getClientRects().length?(e=r.getBoundingClientRect(),n=r.ownerDocument.defaultView,{top:e.top+n.pageYOffset,left:e.left+n.pageXOffset}):{top:0,left:0}:void 0},position:function(){if(this[0]){var e,t,n,r=this[0],i={top:0,left:0};if("fixed"===k.css(r,"position"))t=r.getBoundingClientRect();else{t=this.offset(),n=r.ownerDocument,e=r.offsetParent||n.documentElement;while(e&&(e===n.body||e===n.documentElement)&&"static"===k.css(e,"position"))e=e.parentNode;e&&e!==r&&1===e.nodeType&&((i=k(e).offset()).top+=k.css(e,"borderTopWidth",!0),i.left+=k.css(e,"borderLeftWidth",!0))}return{top:t.top-i.top-k.css(r,"marginTop",!0),left:t.left-i.left-k.css(r,"marginLeft",!0)}}},offsetParent:function(){return this.map(function(){var e=this.offsetParent;while(e&&"static"===k.css(e,"position"))e=e.offsetParent;return e||ie})}}),k.each({scrollLeft:"pageXOffset",scrollTop:"pageYOffset"},function(t,i){var o="pageYOffset"===i;k.fn[t]=function(e){return _(this,function(e,t,n){var r;if(x(e)?r=e:9===e.nodeType&&(r=e.defaultView),void 0===n)return r?r[i]:e[t];r?r.scrollTo(o?r.pageXOffset:n,o?n:r.pageYOffset):e[t]=n},t,e,arguments.length)}}),k.each(["top","left"],function(e,n){k.cssHooks[n]=ze(y.pixelPosition,function(e,t){if(t)return t=_e(e,n),$e.test(t)?k(e).position()[n]+"px":t})}),k.each({Height:"height",Width:"width"},function(a,s){k.each({padding:"inner"+a,content:s,"":"outer"+a},function(r,o){k.fn[o]=function(e,t){var n=arguments.length&&(r||"boolean"!=typeof e),i=r||(!0===e||!0===t?"margin":"border");return _(this,function(e,t,n){var r;return x(e)?0===o.indexOf("outer")?e["inner"+a]:e.document.documentElement["client"+a]:9===e.nodeType?(r=e.documentElement,Math.max(e.body["scroll"+a],r["scroll"+a],e.body["offset"+a],r["offset"+a],r["client"+a])):void 0===n?k.css(e,t,i):k.style(e,t,n,i)},s,n?e:void 0,n)}})}),k.each("blur focus focusin focusout resize scroll click dblclick mousedown mouseup mousemove mouseover mouseout mouseenter mouseleave change select submit keydown keypress keyup contextmenu".split(" "),function(e,n){k.fn[n]=function(e,t){return 0+~]|"+M+")"+M+"*"),U=new RegExp(M+"|>"),X=new RegExp($),V=new RegExp("^"+I+"$"),G={ID:new RegExp("^#("+I+")"),CLASS:new RegExp("^\\.("+I+")"),TAG:new RegExp("^("+I+"|[*])"),ATTR:new RegExp("^"+W),PSEUDO:new RegExp("^"+$),CHILD:new RegExp("^:(only|first|last|nth|nth-last)-(child|of-type)(?:\\("+M+"*(even|odd|(([+-]|)(\\d*)n|)"+M+"*(?:([+-]|)"+M+"*(\\d+)|))"+M+"*\\)|)","i"),bool:new RegExp("^(?:"+R+")$","i"),needsContext:new RegExp("^"+M+"*[>+~]|:(even|odd|eq|gt|lt|nth|first|last)(?:\\("+M+"*((?:-\\d)?\\d*)"+M+"*\\)|)(?=[^-]|$)","i")},Y=/HTML$/i,Q=/^(?:input|select|textarea|button)$/i,J=/^h\d$/i,K=/^[^{]+\{\s*\[native \w/,Z=/^(?:#([\w-]+)|(\w+)|\.([\w-]+))$/,ee=/[+~]/,te=new RegExp("\\\\([\\da-f]{1,6}"+M+"?|("+M+")|.)","ig"),ne=function(e,t,n){var r="0x"+t-65536;return r!=r||n?t:r<0?String.fromCharCode(r+65536):String.fromCharCode(r>>10|55296,1023&r|56320)},re=/([\0-\x1f\x7f]|^-?\d)|^-$|[^\0-\x1f\x7f-\uFFFF\w-]/g,ie=function(e,t){return t?"\0"===e?"\ufffd":e.slice(0,-1)+"\\"+e.charCodeAt(e.length-1).toString(16)+" ":"\\"+e},oe=function(){T()},ae=be(function(e){return!0===e.disabled&&"fieldset"===e.nodeName.toLowerCase()},{dir:"parentNode",next:"legend"});try{H.apply(t=O.call(m.childNodes),m.childNodes),t[m.childNodes.length].nodeType}catch(e){H={apply:t.length?function(e,t){L.apply(e,O.call(t))}:function(e,t){var n=e.length,r=0;while(e[n++]=t[r++]);e.length=n-1}}}function se(t,e,n,r){var i,o,a,s,u,l,c,f=e&&e.ownerDocument,p=e?e.nodeType:9;if(n=n||[],"string"!=typeof t||!t||1!==p&&9!==p&&11!==p)return n;if(!r&&((e?e.ownerDocument||e:m)!==C&&T(e),e=e||C,E)){if(11!==p&&(u=Z.exec(t)))if(i=u[1]){if(9===p){if(!(a=e.getElementById(i)))return n;if(a.id===i)return n.push(a),n}else if(f&&(a=f.getElementById(i))&&y(e,a)&&a.id===i)return n.push(a),n}else{if(u[2])return H.apply(n,e.getElementsByTagName(t)),n;if((i=u[3])&&d.getElementsByClassName&&e.getElementsByClassName)return H.apply(n,e.getElementsByClassName(i)),n}if(d.qsa&&!A[t+" "]&&(!v||!v.test(t))&&(1!==p||"object"!==e.nodeName.toLowerCase())){if(c=t,f=e,1===p&&U.test(t)){(s=e.getAttribute("id"))?s=s.replace(re,ie):e.setAttribute("id",s=k),o=(l=h(t)).length;while(o--)l[o]="#"+s+" "+xe(l[o]);c=l.join(","),f=ee.test(t)&&ye(e.parentNode)||e}try{return H.apply(n,f.querySelectorAll(c)),n}catch(e){A(t,!0)}finally{s===k&&e.removeAttribute("id")}}}return g(t.replace(B,"$1"),e,n,r)}function ue(){var r=[];return function e(t,n){return r.push(t+" ")>b.cacheLength&&delete e[r.shift()],e[t+" "]=n}}function le(e){return e[k]=!0,e}function ce(e){var t=C.createElement("fieldset");try{return!!e(t)}catch(e){return!1}finally{t.parentNode&&t.parentNode.removeChild(t),t=null}}function fe(e,t){var n=e.split("|"),r=n.length;while(r--)b.attrHandle[n[r]]=t}function pe(e,t){var n=t&&e,r=n&&1===e.nodeType&&1===t.nodeType&&e.sourceIndex-t.sourceIndex;if(r)return r;if(n)while(n=n.nextSibling)if(n===t)return-1;return e?1:-1}function de(t){return function(e){return"input"===e.nodeName.toLowerCase()&&e.type===t}}function he(n){return function(e){var t=e.nodeName.toLowerCase();return("input"===t||"button"===t)&&e.type===n}}function ge(t){return function(e){return"form"in e?e.parentNode&&!1===e.disabled?"label"in e?"label"in e.parentNode?e.parentNode.disabled===t:e.disabled===t:e.isDisabled===t||e.isDisabled!==!t&&ae(e)===t:e.disabled===t:"label"in e&&e.disabled===t}}function ve(a){return le(function(o){return o=+o,le(function(e,t){var n,r=a([],e.length,o),i=r.length;while(i--)e[n=r[i]]&&(e[n]=!(t[n]=e[n]))})})}function ye(e){return e&&"undefined"!=typeof e.getElementsByTagName&&e}for(e in d=se.support={},i=se.isXML=function(e){var t=e.namespaceURI,n=(e.ownerDocument||e).documentElement;return!Y.test(t||n&&n.nodeName||"HTML")},T=se.setDocument=function(e){var t,n,r=e?e.ownerDocument||e:m;return r!==C&&9===r.nodeType&&r.documentElement&&(a=(C=r).documentElement,E=!i(C),m!==C&&(n=C.defaultView)&&n.top!==n&&(n.addEventListener?n.addEventListener("unload",oe,!1):n.attachEvent&&n.attachEvent("onunload",oe)),d.attributes=ce(function(e){return e.className="i",!e.getAttribute("className")}),d.getElementsByTagName=ce(function(e){return e.appendChild(C.createComment("")),!e.getElementsByTagName("*").length}),d.getElementsByClassName=K.test(C.getElementsByClassName),d.getById=ce(function(e){return a.appendChild(e).id=k,!C.getElementsByName||!C.getElementsByName(k).length}),d.getById?(b.filter.ID=function(e){var t=e.replace(te,ne);return function(e){return e.getAttribute("id")===t}},b.find.ID=function(e,t){if("undefined"!=typeof t.getElementById&&E){var n=t.getElementById(e);return n?[n]:[]}}):(b.filter.ID=function(e){var n=e.replace(te,ne);return function(e){var t="undefined"!=typeof e.getAttributeNode&&e.getAttributeNode("id");return t&&t.value===n}},b.find.ID=function(e,t){if("undefined"!=typeof t.getElementById&&E){var n,r,i,o=t.getElementById(e);if(o){if((n=o.getAttributeNode("id"))&&n.value===e)return[o];i=t.getElementsByName(e),r=0;while(o=i[r++])if((n=o.getAttributeNode("id"))&&n.value===e)return[o]}return[]}}),b.find.TAG=d.getElementsByTagName?function(e,t){return"undefined"!=typeof t.getElementsByTagName?t.getElementsByTagName(e):d.qsa?t.querySelectorAll(e):void 0}:function(e,t){var n,r=[],i=0,o=t.getElementsByTagName(e);if("*"===e){while(n=o[i++])1===n.nodeType&&r.push(n);return r}return o},b.find.CLASS=d.getElementsByClassName&&function(e,t){if("undefined"!=typeof t.getElementsByClassName&&E)return t.getElementsByClassName(e)},s=[],v=[],(d.qsa=K.test(C.querySelectorAll))&&(ce(function(e){a.appendChild(e).innerHTML="":{dir:"parentNode",first:!0}," ":{dir:"parentNode"},"+":{dir:"previousSibling",first:!0},"~":{dir:"previousSibling"}},preFilter:{ATTR:function(e){return e[1]=e[1].replace(te,ne),e[3]=(e[3]||e[4]||e[5]||"").replace(te,ne),"~="===e[2]&&(e[3]=" "+e[3]+" "),e.slice(0,4)},CHILD:function(e){return e[1]=e[1].toLowerCase(),"nth"===e[1].slice(0,3)?(e[3]||se.error(e[0]),e[4]=+(e[4]?e[5]+(e[6]||1):2*("even"===e[3]||"odd"===e[3])),e[5]=+(e[7]+e[8]||"odd"===e[3])):e[3]&&se.error(e[0]),e},PSEUDO:function(e){var t,n=!e[6]&&e[2];return G.CHILD.test(e[0])?null:(e[3]?e[2]=e[4]||e[5]||"":n&&X.test(n)&&(t=h(n,!0))&&(t=n.indexOf(")",n.length-t)-n.length)&&(e[0]=e[0].slice(0,t),e[2]=n.slice(0,t)),e.slice(0,3))}},filter:{TAG:function(e){var t=e.replace(te,ne).toLowerCase();return"*"===e?function(){return!0}:function(e){return e.nodeName&&e.nodeName.toLowerCase()===t}},CLASS:function(e){var t=p[e+" "];return t||(t=new RegExp("(^|"+M+")"+e+"("+M+"|$)"))&&p(e,function(e){return t.test("string"==typeof e.className&&e.className||"undefined"!=typeof e.getAttribute&&e.getAttribute("class")||"")})},ATTR:function(n,r,i){return function(e){var t=se.attr(e,n);return null==t?"!="===r:!r||(t+="","="===r?t===i:"!="===r?t!==i:"^="===r?i&&0===t.indexOf(i):"*="===r?i&&-1:\x20\t\r\n\f]*)[\x20\t\r\n\f]*\/?>(?:<\/\1>|)$/i;function j(e,n,r){return m(n)?k.grep(e,function(e,t){return!!n.call(e,t,e)!==r}):n.nodeType?k.grep(e,function(e){return e===n!==r}):"string"!=typeof n?k.grep(e,function(e){return-1)[^>]*|#([\w-]+))$/;(k.fn.init=function(e,t,n){var r,i;if(!e)return this;if(n=n||q,"string"==typeof e){if(!(r="<"===e[0]&&">"===e[e.length-1]&&3<=e.length?[null,e,null]:L.exec(e))||!r[1]&&t)return!t||t.jquery?(t||n).find(e):this.constructor(t).find(e);if(r[1]){if(t=t instanceof k?t[0]:t,k.merge(this,k.parseHTML(r[1],t&&t.nodeType?t.ownerDocument||t:E,!0)),D.test(r[1])&&k.isPlainObject(t))for(r in t)m(this[r])?this[r](t[r]):this.attr(r,t[r]);return this}return(i=E.getElementById(r[2]))&&(this[0]=i,this.length=1),this}return e.nodeType?(this[0]=e,this.length=1,this):m(e)?void 0!==n.ready?n.ready(e):e(k):k.makeArray(e,this)}).prototype=k.fn,q=k(E);var H=/^(?:parents|prev(?:Until|All))/,O={children:!0,contents:!0,next:!0,prev:!0};function P(e,t){while((e=e[t])&&1!==e.nodeType);return e}k.fn.extend({has:function(e){var t=k(e,this),n=t.length;return this.filter(function(){for(var e=0;e\x20\t\r\n\f]*)/i,he=/^$|^module$|\/(?:java|ecma)script/i,ge={option:[1,""," "],thead:[1,""],col:[2,""],tr:[2,""],td:[3,""],_default:[0,"",""]};function ve(e,t){var n;return n="undefined"!=typeof e.getElementsByTagName?e.getElementsByTagName(t||"*"):"undefined"!=typeof e.querySelectorAll?e.querySelectorAll(t||"*"):[],void 0===t||t&&A(e,t)?k.merge([e],n):n}function ye(e,t){for(var n=0,r=e.length;nx",y.noCloneChecked=!!me.cloneNode(!0).lastChild.defaultValue;var Te=/^key/,Ce=/^(?:mouse|pointer|contextmenu|drag|drop)|click/,Ee=/^([^.]*)(?:\.(.+)|)/;function ke(){return!0}function Se(){return!1}function Ne(e,t){return e===function(){try{return E.activeElement}catch(e){}}()==("focus"===t)}function Ae(e,t,n,r,i,o){var a,s;if("object"==typeof t){for(s in"string"!=typeof n&&(r=r||n,n=void 0),t)Ae(e,s,n,r,t[s],o);return e}if(null==r&&null==i?(i=n,r=n=void 0):null==i&&("string"==typeof n?(i=r,r=void 0):(i=r,r=n,n=void 0)),!1===i)i=Se;else if(!i)return e;return 1===o&&(a=i,(i=function(e){return k().off(e),a.apply(this,arguments)}).guid=a.guid||(a.guid=k.guid++)),e.each(function(){k.event.add(this,t,i,r,n)})}function De(e,i,o){o?(Q.set(e,i,!1),k.event.add(e,i,{namespace:!1,handler:function(e){var t,n,r=Q.get(this,i);if(1&e.isTrigger&&this[i]){if(r.length)(k.event.special[i]||{}).delegateType&&e.stopPropagation();else if(r=s.call(arguments),Q.set(this,i,r),t=o(this,i),this[i](),r!==(n=Q.get(this,i))||t?Q.set(this,i,!1):n={},r!==n)return e.stopImmediatePropagation(),e.preventDefault(),n.value}else r.length&&(Q.set(this,i,{value:k.event.trigger(k.extend(r[0],k.Event.prototype),r.slice(1),this)}),e.stopImmediatePropagation())}})):void 0===Q.get(e,i)&&k.event.add(e,i,ke)}k.event={global:{},add:function(t,e,n,r,i){var o,a,s,u,l,c,f,p,d,h,g,v=Q.get(t);if(v){n.handler&&(n=(o=n).handler,i=o.selector),i&&k.find.matchesSelector(ie,i),n.guid||(n.guid=k.guid++),(u=v.events)||(u=v.events={}),(a=v.handle)||(a=v.handle=function(e){return"undefined"!=typeof k&&k.event.triggered!==e.type?k.event.dispatch.apply(t,arguments):void 0}),l=(e=(e||"").match(R)||[""]).length;while(l--)d=g=(s=Ee.exec(e[l])||[])[1],h=(s[2]||"").split(".").sort(),d&&(f=k.event.special[d]||{},d=(i?f.delegateType:f.bindType)||d,f=k.event.special[d]||{},c=k.extend({type:d,origType:g,data:r,handler:n,guid:n.guid,selector:i,needsContext:i&&k.expr.match.needsContext.test(i),namespace:h.join(".")},o),(p=u[d])||((p=u[d]=[]).delegateCount=0,f.setup&&!1!==f.setup.call(t,r,h,a)||t.addEventListener&&t.addEventListener(d,a)),f.add&&(f.add.call(t,c),c.handler.guid||(c.handler.guid=n.guid)),i?p.splice(p.delegateCount++,0,c):p.push(c),k.event.global[d]=!0)}},remove:function(e,t,n,r,i){var o,a,s,u,l,c,f,p,d,h,g,v=Q.hasData(e)&&Q.get(e);if(v&&(u=v.events)){l=(t=(t||"").match(R)||[""]).length;while(l--)if(d=g=(s=Ee.exec(t[l])||[])[1],h=(s[2]||"").split(".").sort(),d){f=k.event.special[d]||{},p=u[d=(r?f.delegateType:f.bindType)||d]||[],s=s[2]&&new RegExp("(^|\\.)"+h.join("\\.(?:.*\\.|)")+"(\\.|$)"),a=o=p.length;while(o--)c=p[o],!i&&g!==c.origType||n&&n.guid!==c.guid||s&&!s.test(c.namespace)||r&&r!==c.selector&&("**"!==r||!c.selector)||(p.splice(o,1),c.selector&&p.delegateCount--,f.remove&&f.remove.call(e,c));a&&!p.length&&(f.teardown&&!1!==f.teardown.call(e,h,v.handle)||k.removeEvent(e,d,v.handle),delete u[d])}else for(d in u)k.event.remove(e,d+t[l],n,r,!0);k.isEmptyObject(u)&&Q.remove(e,"handle events")}},dispatch:function(e){var t,n,r,i,o,a,s=k.event.fix(e),u=new Array(arguments.length),l=(Q.get(this,"events")||{})[s.type]||[],c=k.event.special[s.type]||{};for(u[0]=s,t=1;t\x20\t\r\n\f]*)[^>]*)\/>/gi,qe=/\s*$/g;function Oe(e,t){return A(e,"table")&&A(11!==t.nodeType?t:t.firstChild,"tr")&&k(e).children("tbody")[0]||e}function Pe(e){return e.type=(null!==e.getAttribute("type"))+"/"+e.type,e}function Re(e){return"true/"===(e.type||"").slice(0,5)?e.type=e.type.slice(5):e.removeAttribute("type"),e}function Me(e,t){var n,r,i,o,a,s,u,l;if(1===t.nodeType){if(Q.hasData(e)&&(o=Q.access(e),a=Q.set(t,o),l=o.events))for(i in delete a.handle,a.events={},l)for(n=0,r=l[i].length;n")},clone:function(e,t,n){var r,i,o,a,s,u,l,c=e.cloneNode(!0),f=oe(e);if(!(y.noCloneChecked||1!==e.nodeType&&11!==e.nodeType||k.isXMLDoc(e)))for(a=ve(c),r=0,i=(o=ve(e)).length;r").attr(n.scriptAttrs||{}).prop({charset:n.scriptCharset,src:n.url}).on("load error",i=function(e){r.remove(),i=null,e&&t("error"===e.type?404:200,e.type)}),E.head.appendChild(r[0])},abort:function(){i&&i()}}});var Vt,Gt=[],Yt=/(=)\?(?=&|$)|\?\?/;k.ajaxSetup({jsonp:"callback",jsonpCallback:function(){var e=Gt.pop()||k.expando+"_"+kt++;return this[e]=!0,e}}),k.ajaxPrefilter("json jsonp",function(e,t,n){var r,i,o,a=!1!==e.jsonp&&(Yt.test(e.url)?"url":"string"==typeof e.data&&0===(e.contentType||"").indexOf("application/x-www-form-urlencoded")&&Yt.test(e.data)&&"data");if(a||"jsonp"===e.dataTypes[0])return r=e.jsonpCallback=m(e.jsonpCallback)?e.jsonpCallback():e.jsonpCallback,a?e[a]=e[a].replace(Yt,"$1"+r):!1!==e.jsonp&&(e.url+=(St.test(e.url)?"&":"?")+e.jsonp+"="+r),e.converters["script json"]=function(){return o||k.error(r+" was not called"),o[0]},e.dataTypes[0]="json",i=C[r],C[r]=function(){o=arguments},n.always(function(){void 0===i?k(C).removeProp(r):C[r]=i,e[r]&&(e.jsonpCallback=t.jsonpCallback,Gt.push(r)),o&&m(i)&&i(o[0]),o=i=void 0}),"script"}),y.createHTMLDocument=((Vt=E.implementation.createHTMLDocument("").body).innerHTML="",2===Vt.childNodes.length),k.parseHTML=function(e,t,n){return"string"!=typeof e?[]:("boolean"==typeof t&&(n=t,t=!1),t||(y.createHTMLDocument?((r=(t=E.implementation.createHTMLDocument("")).createElement("base")).href=E.location.href,t.head.appendChild(r)):t=E),o=!n&&[],(i=D.exec(e))?[t.createElement(i[1])]:(i=we([e],t,o),o&&o.length&&k(o).remove(),k.merge([],i.childNodes)));var r,i,o},k.fn.load=function(e,t,n){var r,i,o,a=this,s=e.indexOf(" ");return-1").append(k.parseHTML(e)).find(r):e)}).always(n&&function(e,t){a.each(function(){n.apply(this,o||[e.responseText,t,e])})}),this},k.each(["ajaxStart","ajaxStop","ajaxComplete","ajaxError","ajaxSuccess","ajaxSend"],function(e,t){k.fn[t]=function(e){return this.on(t,e)}}),k.expr.pseudos.animated=function(t){return k.grep(k.timers,function(e){return t===e.elem}).length},k.offset={setOffset:function(e,t,n){var r,i,o,a,s,u,l=k.css(e,"position"),c=k(e),f={};"static"===l&&(e.style.position="relative"),s=c.offset(),o=k.css(e,"top"),u=k.css(e,"left"),("absolute"===l||"fixed"===l)&&-1<(o+u).indexOf("auto")?(a=(r=c.position()).top,i=r.left):(a=parseFloat(o)||0,i=parseFloat(u)||0),m(t)&&(t=t.call(e,n,k.extend({},s))),null!=t.top&&(f.top=t.top-s.top+a),null!=t.left&&(f.left=t.left-s.left+i),"using"in t?t.using.call(e,f):c.css(f)}},k.fn.extend({offset:function(t){if(arguments.length)return void 0===t?this:this.each(function(e){k.offset.setOffset(this,t,e)});var e,n,r=this[0];return r?r.getClientRects().length?(e=r.getBoundingClientRect(),n=r.ownerDocument.defaultView,{top:e.top+n.pageYOffset,left:e.left+n.pageXOffset}):{top:0,left:0}:void 0},position:function(){if(this[0]){var e,t,n,r=this[0],i={top:0,left:0};if("fixed"===k.css(r,"position"))t=r.getBoundingClientRect();else{t=this.offset(),n=r.ownerDocument,e=r.offsetParent||n.documentElement;while(e&&(e===n.body||e===n.documentElement)&&"static"===k.css(e,"position"))e=e.parentNode;e&&e!==r&&1===e.nodeType&&((i=k(e).offset()).top+=k.css(e,"borderTopWidth",!0),i.left+=k.css(e,"borderLeftWidth",!0))}return{top:t.top-i.top-k.css(r,"marginTop",!0),left:t.left-i.left-k.css(r,"marginLeft",!0)}}},offsetParent:function(){return this.map(function(){var e=this.offsetParent;while(e&&"static"===k.css(e,"position"))e=e.offsetParent;return e||ie})}}),k.each({scrollLeft:"pageXOffset",scrollTop:"pageYOffset"},function(t,i){var o="pageYOffset"===i;k.fn[t]=function(e){return _(this,function(e,t,n){var r;if(x(e)?r=e:9===e.nodeType&&(r=e.defaultView),void 0===n)return r?r[i]:e[t];r?r.scrollTo(o?r.pageXOffset:n,o?n:r.pageYOffset):e[t]=n},t,e,arguments.length)}}),k.each(["top","left"],function(e,n){k.cssHooks[n]=ze(y.pixelPosition,function(e,t){if(t)return t=_e(e,n),$e.test(t)?k(e).position()[n]+"px":t})}),k.each({Height:"height",Width:"width"},function(a,s){k.each({padding:"inner"+a,content:s,"":"outer"+a},function(r,o){k.fn[o]=function(e,t){var n=arguments.length&&(r||"boolean"!=typeof e),i=r||(!0===e||!0===t?"margin":"border");return _(this,function(e,t,n){var r;return x(e)?0===o.indexOf("outer")?e["inner"+a]:e.document.documentElement["client"+a]:9===e.nodeType?(r=e.documentElement,Math.max(e.body["scroll"+a],r["scroll"+a],e.body["offset"+a],r["offset"+a],r["client"+a])):void 0===n?k.css(e,t,i):k.style(e,t,n,i)},s,n?e:void 0,n)}})}),k.each("blur focus focusin focusout resize scroll click dblclick mousedown mouseup mousemove mouseover mouseout mouseenter mouseleave change select submit keydown keypress keyup contextmenu".split(" "),function(e,n){k.fn[n]=function(e,t){return 0 1)):
+ elif (
+ (ctx.triggered[0]["prop_id"] == "prediction_picking.selectedData")
+ and (selected_data is not None)
+ and (len(selected_data) > 1)
+ ):
df = select_data_from_prediction_picking(self.round_dataframe, selected_data)
# If click on reset button
- elif ctx.triggered[0]['prop_id'] == 'reset_dropdown_button.n_clicks':
+ elif ctx.triggered[0]["prop_id"] == "reset_dropdown_button.n_clicks":
df = self.round_dataframe
# If click on Apply filter
- elif ((ctx.triggered[0]['prop_id'] == 'apply_filter.n_clicks') | (
- (ctx.triggered[0]['prop_id'] == 'prediction_picking.selectedData') and
- ((selected_data is None))) | (
- (ctx.triggered[0]['prop_id'] == 'prediction_picking.selectedData') and
- (selected_data is not None and len(selected_data) == 1 and
- len(selected_data['points'])!=0 and selected_data['points'][0]['curveNumber'] > 0)
- )):
+ elif (
+ (ctx.triggered[0]["prop_id"] == "apply_filter.n_clicks")
+ | ((ctx.triggered[0]["prop_id"] == "prediction_picking.selectedData") and (selected_data is None))
+ | (
+ (ctx.triggered[0]["prop_id"] == "prediction_picking.selectedData")
+ and (
+ selected_data is not None
+ and len(selected_data) == 1
+ and len(selected_data["points"]) != 0
+ and selected_data["points"][0]["curveNumber"] > 0
+ )
+ )
+ ):
df = self.round_dataframe.copy()
- feature_id = [id_feature[i]['index'] for i in range(len(id_feature))]
+ feature_id = [id_feature[i]["index"] for i in range(len(id_feature))]
df = select_data_from_str_filters(
df,
- feature_id,
- id_str_modality,
- val_feature,
+ feature_id,
+ id_str_modality,
+ val_feature,
val_str_modality,
)
df = select_data_from_bool_filters(
@@ -1500,25 +1588,25 @@ def update_datatable(selected_data,
df = select_data_from_date_filters(
df,
feature_id,
- id_date,
+ id_date,
val_feature,
start_date,
end_date,
)
- filtered_subset_info = f"Subset length: {len(df)} ({int(round(100*len(df)/self.explainer.x_init.shape[0]))}%)"
- if len(df)==0:
- filtered_subset_color="danger"
-
+ filtered_subset_info = (
+ f"Subset length: {len(df)} ({int(round(100*len(df)/self.explainer.x_init.shape[0]))}%)"
+ )
+ if len(df) == 0:
+ filtered_subset_color = "danger"
+
elif None not in nclicks_del:
df = self.round_dataframe
else:
raise dash.exceptions.PreventUpdate
- data = df.to_dict('records')
+ data = df.to_dict("records")
tooltip_data = [
- {
- column: {'value': str(value), 'type': 'text'}
- for column, value in row.items()
- } for row in df.to_dict('index').values()
+ {column: {"value": str(value), "type": "text"} for column, value in row.items()}
+ for row in df.to_dict("index").values()
]
return (
data,
@@ -1530,41 +1618,43 @@ def update_datatable(selected_data,
@app.callback(
[
- Output('global_feature_importance', 'figure'),
- Output('global_feature_importance', 'clickData'),
- Output('clickdata-store', 'data')
+ Output("global_feature_importance", "figure"),
+ Output("global_feature_importance", "clickData"),
+ Output("clickdata-store", "data"),
],
[
- Input('select_label', 'value'),
- Input('dataset', 'data'),
- Input('prediction_picking', 'selectedData'),
- Input('apply_filter', 'n_clicks'),
- Input('reset_dropdown_button', 'n_clicks'),
- Input({'type': 'del_dropdown_button', 'index': ALL}, 'n_clicks'),
- Input('card_global_feature_importance', 'n_clicks'),
- Input('bool_groups', 'on'),
- Input('ember_global_feature_importance', 'n_clicks')
+ Input("select_label", "value"),
+ Input("dataset", "data"),
+ Input("prediction_picking", "selectedData"),
+ Input("apply_filter", "n_clicks"),
+ Input("reset_dropdown_button", "n_clicks"),
+ Input({"type": "del_dropdown_button", "index": ALL}, "n_clicks"),
+ Input("card_global_feature_importance", "n_clicks"),
+ Input("bool_groups", "on"),
+ Input("ember_global_feature_importance", "n_clicks"),
],
[
- State('global_feature_importance', 'clickData'),
- State('global_feature_importance', 'figure'),
- State('features', 'value'),
- State('clickdata-store', 'data')
- ]
+ State("global_feature_importance", "clickData"),
+ State("global_feature_importance", "figure"),
+ State("features", "value"),
+ State("clickdata-store", "data"),
+ ],
)
- def update_feature_importance(label,
- data,
- selected_data,
- apply_filters,
- reset_filter,
- nclicks_del,
- n_clicks,
- bool_group,
- click_zoom,
- clickData,
- figure,
- features,
- clickData_store):
+ def update_feature_importance(
+ label,
+ data,
+ selected_data,
+ apply_filters,
+ reset_filter,
+ nclicks_del,
+ n_clicks,
+ bool_group,
+ click_zoom,
+ clickData,
+ figure,
+ features,
+ clickData_store,
+ ):
"""
update feature importance plot according label, click on graph,
filters applied and subset selected in prediction picking graph.
@@ -1593,29 +1683,32 @@ def update_feature_importance(label,
zoom_active = get_figure_zoom(click_zoom)
selection = None
list_index = self.list_index
- if clickData is not None and (ctx.triggered[0]['prop_id'] in [
- 'apply_filter.n_clicks',
- 'reset_dropdown_button.n_clicks',
- 'dataset.data'
- ] or ('del_dropdown_button' in ctx.triggered[0]['prop_id'] and
- None not in nclicks_del)):
+ if clickData is not None and (
+ ctx.triggered[0]["prop_id"]
+ in ["apply_filter.n_clicks", "reset_dropdown_button.n_clicks", "dataset.data"]
+ or ("del_dropdown_button" in ctx.triggered[0]["prop_id"] and None not in nclicks_del)
+ ):
clickData = update_click_data_on_subset_changes(clickData)
- selected_feature = self.explainer.inv_features_dict.get(
- get_feature_from_clicked_data(clickData)
- ) if clickData else None
+ selected_feature = (
+ self.explainer.inv_features_dict.get(get_feature_from_clicked_data(clickData)) if clickData else None
+ )
if self.explainer.features_groups and bool_group:
- if ctx.triggered[0]['prop_id'] == 'card_global_feature_importance.n_clicks':
+ if ctx.triggered[0]["prop_id"] == "card_global_feature_importance.n_clicks":
# When we click twice on the same bar this will reset the graph
if clickData_store == clickData:
selected_feature = None
- selected_feature = get_feature_from_features_groups(selected_feature, self.explainer.features_groups)
- elif ctx.triggered[0]['prop_id'] == 'ember_global_feature_importance.n_clicks':
- selected_feature = get_feature_from_features_groups(selected_feature, self.explainer.features_groups)
+ selected_feature = get_feature_from_features_groups(
+ selected_feature, self.explainer.features_groups
+ )
+ elif ctx.triggered[0]["prop_id"] == "ember_global_feature_importance.n_clicks":
+ selected_feature = get_feature_from_features_groups(
+ selected_feature, self.explainer.features_groups
+ )
else:
selected_feature = None
-
+
selection = get_indexes_from_datatable(data, list_index)
group_name = get_group_name(selected_feature, self.explainer.features_groups)
@@ -1626,44 +1719,35 @@ def update_feature_importance(label,
label=label,
group_name=group_name,
display_groups=bool_group,
- zoom=zoom_active
+ zoom=zoom_active,
)
# Adjust graph with adding x axis title
- MyGraph.adjust_graph_static(figure, x_ax='Mean absolute Contribution')
- figure.layout.clickmode = 'event+select'
+ MyGraph.adjust_graph_static(figure, x_ax="Mean absolute Contribution")
+ figure.layout.clickmode = "event+select"
if selected_feature:
self.select_point(figure, clickData)
# font size can be adapted to screen size
- nb_car = max([len(figure.data[0].y[i]) for i in
- range(len(figure.data[0].y))])
- figure.update_layout(
- yaxis=dict(tickfont={'size': min(round(500 / nb_car), 12)})
- )
+ nb_car = max([len(figure.data[0].y[i]) for i in range(len(figure.data[0].y))])
+ figure.update_layout(yaxis=dict(tickfont={"size": min(round(500 / nb_car), 12)}))
clickData_store = clickData.copy() if clickData is not None else None
return figure, clickData, clickData_store
@app.callback(
- Output(component_id='feature_selector', component_property='figure'),
+ Output(component_id="feature_selector", component_property="figure"),
[
- Input('global_feature_importance', 'clickData'),
- Input('dataset', 'data'),
- Input('select_label', 'value'),
- Input('ember_feature_selector', 'n_clicks')
+ Input("global_feature_importance", "clickData"),
+ Input("dataset", "data"),
+ Input("select_label", "value"),
+ Input("ember_feature_selector", "n_clicks"),
],
[
- State('points', 'value'),
- State('violin', 'value'),
- State('global_feature_importance', 'figure'),
- ]
+ State("points", "value"),
+ State("violin", "value"),
+ State("global_feature_importance", "figure"),
+ ],
)
- def update_feature_selector(feature,
- data,
- label,
- click_zoom,
- points,
- violin,
- gfi_figure):
+ def update_feature_selector(feature, data, label, click_zoom, points, violin, gfi_figure):
"""
Update feature plot according to label, data,
selected feature on features importance graph,
@@ -1689,55 +1773,45 @@ def update_feature_selector(feature,
else:
selected_feature = self.selected_feature
- if feature is not None and feature['points'][0]['curveNumber'] == 0 and \
- len(gfi_figure['data']) == 2:
+ if feature is not None and feature["points"][0]["curveNumber"] == 0 and len(gfi_figure["data"]) == 2:
subset = get_indexes_from_datatable(data, list_index)
else:
subset = None
fs_figure = self.explainer.plot.contribution_plot(
- col=selected_feature,
- selection=subset,
- label=label,
- violin_maxf=violin,
- max_points=points,
- zoom=zoom_active
- )
+ col=selected_feature,
+ selection=subset,
+ label=label,
+ violin_maxf=violin,
+ max_points=points,
+ zoom=zoom_active,
+ )
- fs_figure['layout'].clickmode = 'event+select'
+ fs_figure["layout"].clickmode = "event+select"
# Adjust graph with adding x and y axis titles
- MyGraph.adjust_graph_static(fs_figure,
- #x_ax=truncate_str(self.layout.selected_feature, 110),
+ MyGraph.adjust_graph_static(
+ fs_figure,
+ # x_ax=truncate_str(self.layout.selected_feature, 110),
x_ax=truncate_str(selected_feature, 110),
- y_ax='Contribution')
+ y_ax="Contribution",
+ )
return fs_figure
@app.callback(
+ [Output("index_id", "value"), Output("index_id", "n_submit")],
[
- Output('index_id', 'value'),
- Output("index_id", "n_submit")
+ Input("feature_selector", "clickData"),
+ Input("prediction_picking", "clickData"),
+ Input("dataset", "active_cell"),
+ Input("apply_filter", "n_clicks"),
+ Input("reset_dropdown_button", "n_clicks"),
+ Input({"type": "del_dropdown_button", "index": ALL}, "n_clicks"),
],
- [
- Input('feature_selector', 'clickData'),
- Input('prediction_picking', 'clickData'),
- Input('dataset', 'active_cell'),
- Input('apply_filter', 'n_clicks'),
- Input('reset_dropdown_button', 'n_clicks'),
- Input({'type': 'del_dropdown_button', 'index': ALL}, 'n_clicks')
- ],
- [
- State('dataset', 'data'),
- State('index_id', 'value') # Get the current value of the index
- ]
+ [State("dataset", "data"), State("index_id", "value")], # Get the current value of the index
)
- def update_index_id(click_data,
- prediction_picking,
- cell,
- apply_filters,
- reset_filter,
- nclicks_del,
- data,
- current_index_id):
+ def update_index_id(
+ click_data, prediction_picking, cell, apply_filters, reset_filter, nclicks_del, data, current_index_id
+ ):
"""
This function is used to update index value according to
active cell, filters and click data on feature plot or on
@@ -1758,91 +1832,75 @@ def update_index_id(click_data,
"""
ctx = dash.callback_context
selected = None
- if ctx.triggered[0]['prop_id'] == 'feature_selector.clickData':
- selected = click_data['points'][0]['customdata'][1]
- elif ctx.triggered[0]['prop_id'] == 'prediction_picking.clickData':
- selected = prediction_picking['points'][0]['customdata']
- elif ctx.triggered[0]['prop_id'] == 'dataset.active_cell':
- selected = data[cell['row']]['_index_']
- elif (('del_dropdown_button' in ctx.triggered[0]['prop_id']) &
- (None in nclicks_del)):
+ if ctx.triggered[0]["prop_id"] == "feature_selector.clickData":
+ selected = click_data["points"][0]["customdata"][1]
+ elif ctx.triggered[0]["prop_id"] == "prediction_picking.clickData":
+ selected = prediction_picking["points"][0]["customdata"]
+ elif ctx.triggered[0]["prop_id"] == "dataset.active_cell":
+ selected = data[cell["row"]]["_index_"]
+ elif ("del_dropdown_button" in ctx.triggered[0]["prop_id"]) & (None in nclicks_del):
selected = current_index_id
return selected, True
- @app.callback(
- Output('threshold_label', 'children'),
- [Input('threshold_id', 'value')])
+ @app.callback(Output("threshold_label", "children"), [Input("threshold_id", "value")])
def update_threshold_label(value):
"""
update threshold label
"""
- return f'Threshold: {value}'
+ return f"Threshold: {value}"
- @app.callback(
- Output('max_contrib_label', 'children'),
- [Input('max_contrib_id', 'value')])
+ @app.callback(Output("max_contrib_label", "children"), [Input("max_contrib_id", "value")])
def update_max_contrib_label(value):
"""
update max_contrib label
"""
- return f'Features to display: {value}'
+ return f"Features to display: {value}"
@app.callback(
- [Output('max_contrib_id', 'value'),
- Output('max_contrib_id', 'max'),
- Output('max_contrib_id', 'marks')
- ],
- [Input('modal', 'is_open')],
- [State('features', 'value'),
- State('max_contrib_id', 'value')]
+ [Output("max_contrib_id", "value"), Output("max_contrib_id", "max"), Output("max_contrib_id", "marks")],
+ [Input("modal", "is_open")],
+ [State("features", "value"), State("max_contrib_id", "value")],
)
- def update_max_contrib_id(is_open,
- features,
- value):
+ def update_max_contrib_id(is_open, features, value):
"""
update max contrib component layout after settings modifications
"""
ctx = dash.callback_context
- if ctx.triggered[0]['prop_id'] == 'modal.is_open':
+ if ctx.triggered[0]["prop_id"] == "modal.is_open":
if is_open:
raise PreventUpdate
else:
- value, max, marks = update_features_to_display(
- features,
- len(self.explainer.x_init.columns),
- value
- )
+ value, max, marks = update_features_to_display(features, len(self.explainer.x_init.columns), value)
return value, max, marks
@app.callback(
- Output(component_id='detail_feature', component_property='figure'),
+ Output(component_id="detail_feature", component_property="figure"),
[
- Input('threshold_id', 'value'),
- Input('max_contrib_id', 'value'),
- Input('check_id_positive', 'value'),
- Input('check_id_negative', 'value'),
- Input('masked_contrib_id', 'value'),
- Input('select_label', 'value'),
+ Input("threshold_id", "value"),
+ Input("max_contrib_id", "value"),
+ Input("check_id_positive", "value"),
+ Input("check_id_negative", "value"),
+ Input("masked_contrib_id", "value"),
+ Input("select_label", "value"),
Input("validation", "n_clicks"),
- Input('bool_groups', 'on'),
- Input('ember_detail_feature', 'n_clicks'),
+ Input("bool_groups", "on"),
+ Input("ember_detail_feature", "n_clicks"),
],
- [
- State('index_id', 'value'),
- State('dataset', 'data')
- ]
+ [State("index_id", "value"), State("dataset", "data")],
)
- def update_detail_feature(threshold,
- max_contrib,
- positive,
- negative,
- masked,
- label,
- validation_click,
- bool_group,
- click_zoom,
- index,
- data):
+ def update_detail_feature(
+ threshold,
+ max_contrib,
+ positive,
+ negative,
+ masked,
+ label,
+ validation_click,
+ bool_group,
+ click_zoom,
+ index,
+ data,
+ ):
"""
update local explanation plot according to app changes.
-------------------------------------------------------
@@ -1871,39 +1929,36 @@ def update_detail_feature(threshold,
selected = None
threshold = threshold if threshold != 0 else None
sign = get_feature_contributions_sign_to_show(positive, negative)
- self.explainer.filter(threshold=threshold,
- features_to_hide=masked,
- positive=sign,
- max_contrib=max_contrib,
- display_groups=bool_group)
+ self.explainer.filter(
+ threshold=threshold,
+ features_to_hide=masked,
+ positive=sign,
+ max_contrib=max_contrib,
+ display_groups=bool_group,
+ )
figure = self.explainer.plot.local_plot(
index=selected,
label=label,
show_masked=True,
yaxis_max_label=8,
display_groups=bool_group,
- zoom=zoom_active
+ zoom=zoom_active,
)
if selected is not None:
# Adjust graph with adding x axis titles
- MyGraph.adjust_graph_static(figure, x_ax='Contribution')
+ MyGraph.adjust_graph_static(figure, x_ax="Contribution")
# font size can be adapted to screen size
- list_yaxis = [figure.data[i].y[0] for i in
- range(len(figure.data))]
+ list_yaxis = [figure.data[i].y[0] for i in range(len(figure.data))]
# exclude new line with labels of y axis
if list_yaxis != []:
- list_yaxis = [x.split('' + new_title + ' '
- }
+ "text": '' + new_title + " ",
+ },
)
# update x title and font-size of the title
- figure.update_xaxes(title='' + x_ax + ' ',
- automargin=True
- )
+ figure.update_xaxes(title='' + x_ax + " ", automargin=True)
# update y title and font-size of the title
- figure.update_yaxes(title='' + y_ax + ' ',
- automargin=True
- )
+ figure.update_yaxes(title='' + y_ax + " ", automargin=True)
def update_title(title):
@@ -96,10 +88,10 @@ def update_title(title):
-------
str
"""
- patt = re.compile('^(.+)$')
+ patt = re.compile("^(.+)$")
try:
- list_non_empty_str_matches = [x for x in patt.findall(title)[0] if x != '']
- updated = ' - '.join(map(str, list_non_empty_str_matches))
+ list_non_empty_str_matches = [x for x in patt.findall(title)[0] if x != ""]
+ updated = " - ".join(map(str, list_non_empty_str_matches))
except:
updated = title
return updated
diff --git a/shapash/webapp/utils/callbacks.py b/shapash/webapp/utils/callbacks.py
index 3275e29c..c5748e4a 100644
--- a/shapash/webapp/utils/callbacks.py
+++ b/shapash/webapp/utils/callbacks.py
@@ -1,10 +1,10 @@
+import datetime
from typing import Optional, Tuple
+import dash_bootstrap_components as dbc
import numpy as np
import pandas as pd
-import datetime
from dash import dcc, html
-import dash_bootstrap_components as dbc
def select_data_from_prediction_picking(round_dataframe: pd.DataFrame, selected_data: dict) -> pd.DataFrame:
@@ -23,18 +23,18 @@ def select_data_from_prediction_picking(round_dataframe: pd.DataFrame, selected_
Subset dataframe
"""
row_ids = []
- for p in selected_data['points']:
- row_ids.append(p['customdata'])
+ for p in selected_data["points"]:
+ row_ids.append(p["customdata"])
df = round_dataframe.loc[row_ids]
-
+
return df
def select_data_from_str_filters(
df: pd.DataFrame,
- feature_id: list,
- id_str_modality: list,
- val_feature: list,
+ feature_id: list,
+ id_str_modality: list,
+ val_feature: list,
val_str_modality: list,
) -> pd.DataFrame:
"""Create a subset dataframe from filters.
@@ -58,23 +58,23 @@ def select_data_from_str_filters(
Subset dataframe
"""
# get list of ID
- str_id = [id_str_modality[i]['index'] for i in range(len(id_str_modality))]
+ str_id = [id_str_modality[i]["index"] for i in range(len(id_str_modality))]
# If there is some filters
if len(str_id) > 0:
for i in range(len(feature_id)):
if feature_id[i] in str_id:
position = np.where(np.array(str_id) == feature_id[i])[0][0]
- if ((position is not None) & (val_str_modality[position] is not None)):
+ if (position is not None) & (val_str_modality[position] is not None):
df = df[df[val_feature[i]].isin(val_str_modality[position])]
-
+
return df
def select_data_from_bool_filters(
df: pd.DataFrame,
- feature_id: list,
- id_bool_modality: list,
- val_feature: list,
+ feature_id: list,
+ id_bool_modality: list,
+ val_feature: list,
val_bool_modality: list,
) -> pd.DataFrame:
"""Create a subset dataframe from filters.
@@ -98,23 +98,23 @@ def select_data_from_bool_filters(
Subset dataframe
"""
# get list of ID
- bool_id = [id_bool_modality[i]['index'] for i in range(len(id_bool_modality))]
+ bool_id = [id_bool_modality[i]["index"] for i in range(len(id_bool_modality))]
# If there is some filters
if len(bool_id) > 0:
for i in range(len(feature_id)):
if feature_id[i] in bool_id:
position = np.where(np.array(bool_id) == feature_id[i])[0][0]
- if ((position is not None) & (val_bool_modality[position] is not None)):
+ if (position is not None) & (val_bool_modality[position] is not None):
df = df[df[val_feature[i]] == val_bool_modality[position]]
-
+
return df
def select_data_from_numeric_filters(
df: pd.DataFrame,
- feature_id: list,
- id_lower_modality: list,
- val_feature: list,
+ feature_id: list,
+ id_lower_modality: list,
+ val_feature: list,
val_lower_modality: list,
val_upper_modality: list,
) -> pd.DataFrame:
@@ -141,26 +141,31 @@ def select_data_from_numeric_filters(
Subset dataframe
"""
# get list of ID
- lower_id = [id_lower_modality[i]['index'] for i in range(len(id_lower_modality))]
+ lower_id = [id_lower_modality[i]["index"] for i in range(len(id_lower_modality))]
# If there is some filters
if len(lower_id) > 0:
for i in range(len(feature_id)):
if feature_id[i] in lower_id:
position = np.where(np.array(lower_id) == feature_id[i])[0][0]
- if((position is not None) & (val_lower_modality[position] is not None) &
- (val_upper_modality[position] is not None)):
- if (val_lower_modality[position] < val_upper_modality[position]):
- df = df[(df[val_feature[i]] >= val_lower_modality[position]) &
- (df[val_feature[i]] <= val_upper_modality[position])]
-
+ if (
+ (position is not None)
+ & (val_lower_modality[position] is not None)
+ & (val_upper_modality[position] is not None)
+ ):
+ if val_lower_modality[position] < val_upper_modality[position]:
+ df = df[
+ (df[val_feature[i]] >= val_lower_modality[position])
+ & (df[val_feature[i]] <= val_upper_modality[position])
+ ]
+
return df
def select_data_from_date_filters(
df: pd.DataFrame,
- feature_id: list,
- id_date: list,
- val_feature: list,
+ feature_id: list,
+ id_date: list,
+ val_feature: list,
start_date: list,
end_date: list,
) -> pd.DataFrame:
@@ -187,17 +192,15 @@ def select_data_from_date_filters(
Subset dataframe
"""
# get list of ID
- date_id = [id_date[i]['index'] for i in range(len(id_date))]
+ date_id = [id_date[i]["index"] for i in range(len(id_date))]
# If there is some filters
if len(date_id) > 0:
for i in range(len(feature_id)):
if feature_id[i] in date_id:
position = np.where(np.array(date_id) == feature_id[i])[0][0]
- if((position is not None) &
- (start_date[position] < end_date[position])):
- df = df[((df[val_feature[i]] >= start_date[position]) &
- (df[val_feature[i]] <= end_date[position]))]
-
+ if (position is not None) & (start_date[position] < end_date[position]):
+ df = df[((df[val_feature[i]] >= start_date[position]) & (df[val_feature[i]] <= end_date[position]))]
+
return df
@@ -214,7 +217,7 @@ def get_feature_from_clicked_data(click_data: dict) -> str:
str
Selected feature
"""
- selected_feature = click_data['points'][0]['label'].replace('', '').replace(' ', '')
+ selected_feature = click_data["points"][0]["label"].replace("", "").replace(" ", "")
return selected_feature
@@ -233,8 +236,7 @@ def get_feature_from_features_groups(selected_feature: Optional[str], features_g
Optional[str]
Group feature name or selected feature if not in a group.
"""
- list_sub_features = [f for group_features in features_groups.values()
- for f in group_features]
+ list_sub_features = [f for group_features in features_groups.values() for f in group_features]
if selected_feature in list_sub_features:
for k, v in features_groups.items():
if selected_feature in v:
@@ -257,14 +259,14 @@ def get_group_name(selected_feature: Optional[str], features_groups: Optional[di
Optional[str]
Group feature name
"""
- group_name = selected_feature if (
- features_groups is not None and selected_feature in features_groups.keys()
- ) else None
+ group_name = (
+ selected_feature if (features_groups is not None and selected_feature in features_groups.keys()) else None
+ )
return group_name
def get_indexes_from_datatable(data: list, list_index: Optional[list] = None) -> Optional[list]:
- """Get the indexes of the data. If list_index is given and is the same length than
+ """Get the indexes of the data. If list_index is given and is the same length than
the indexes, there is no subset selected.
Parameters
@@ -279,8 +281,8 @@ def get_indexes_from_datatable(data: list, list_index: Optional[list] = None) ->
Optional[list]
Indexes of the data
"""
- indexes = [d['_index_'] for d in data]
- if list_index is not None and (len(indexes)==len(list_index) or len(indexes)==0):
+ indexes = [d["_index_"] for d in data]
+ if list_index is not None and (len(indexes) == len(list_index) or len(indexes) == 0):
indexes = None
return indexes
@@ -298,13 +300,13 @@ def update_click_data_on_subset_changes(click_data: dict) -> dict:
dict
Updated feature importance click data
"""
- point = click_data['points'][0]
- point['curveNumber'] = 0
- click_data = {'points':[point]}
+ point = click_data["points"][0]
+ point["curveNumber"] = 0
+ click_data = {"points": [point]}
return click_data
-def get_figure_zoom(click_zoom: int) -> bool :
+def get_figure_zoom(click_zoom: int) -> bool:
"""Get figure zoom from n_clicks
Parameters
@@ -341,9 +343,9 @@ def get_feature_contributions_sign_to_show(positive: list, negative: list) -> Op
Sign to show on plot
"""
if positive == [1]:
- sign = (None if negative == [1] else True)
+ sign = None if negative == [1] else True
else:
- sign = (False if negative == [1] else None)
+ sign = False if negative == [1] else None
return sign
@@ -375,9 +377,11 @@ def update_features_to_display(features: int, nb_columns: int, value: int) -> Tu
nb_marks = min(int(max_value // 7), 10)
else:
nb_marks = 2
- marks = {f'{round(max_value * feat / nb_marks)}': f'{round(max_value * feat / nb_marks)}'
- for feat in range(1, nb_marks + 1)}
- marks['1'] = '1'
+ marks = {
+ f"{round(max_value * feat / nb_marks)}": f"{round(max_value * feat / nb_marks)}"
+ for feat in range(1, nb_marks + 1)
+ }
+ marks["1"] = "1"
if max_value < value:
value = max_value
@@ -404,18 +408,12 @@ def get_id_card_features(data: list, selected: int, special_cols: list, features
Dataframe of the features
"""
selected_row = pd.DataFrame([data[selected]], index=["feature_value"]).T
- selected_row["feature_name"] = selected_row.index.map(
- lambda x: x if x in special_cols else features_dict[x]
- )
+ selected_row["feature_name"] = selected_row.index.map(lambda x: x if x in special_cols else features_dict[x])
return selected_row
def get_id_card_contrib(
- data: dict,
- index: int,
- features_dict: dict,
- columns_dict: dict,
- label_num: int = None
+ data: dict, index: int, features_dict: dict, columns_dict: dict, label_num: int = None
) -> pd.DataFrame:
"""Get the contributions of the selected index for the identity card.
@@ -438,11 +436,11 @@ def get_id_card_contrib(
Dataframe of the contributions
"""
if label_num is not None:
- contrib = data['contrib_sorted'][label_num].loc[index, :].values
- var_dict = data['var_dict'][label_num].loc[index, :].values
+ contrib = data["contrib_sorted"][label_num].loc[index, :].values
+ var_dict = data["var_dict"][label_num].loc[index, :].values
else:
- contrib = data['contrib_sorted'].loc[index, :].values
- var_dict = data['var_dict'].loc[index, :].values
+ contrib = data["contrib_sorted"].loc[index, :].values
+ var_dict = data["var_dict"].loc[index, :].values
var_dict = [features_dict[columns_dict[x]] for x in var_dict]
selected_contrib = pd.DataFrame([var_dict, contrib], index=["feature_name", "feature_contrib"]).T
@@ -453,11 +451,11 @@ def get_id_card_contrib(
def create_id_card_data(
selected_row: pd.DataFrame,
- selected_contrib: pd.DataFrame,
- sort_by: str,
+ selected_contrib: pd.DataFrame,
+ sort_by: str,
order: bool,
- special_cols: list,
- additional_features_dict: dict
+ special_cols: list,
+ additional_features_dict: dict,
) -> pd.DataFrame:
"""Merge and sort features and contributions dataframes for the identity card.
@@ -483,11 +481,15 @@ def create_id_card_data(
"""
selected_data = selected_row.merge(selected_contrib, how="left", on="feature_name")
selected_data.index = selected_row.index
- selected_data = pd.concat([
- selected_data.loc[special_cols],
- selected_data.drop(index=special_cols+list(additional_features_dict.keys())).sort_values(sort_by, ascending=order),
- selected_data.loc[list(additional_features_dict.keys())].sort_values(sort_by, ascending=order)
- ])
+ selected_data = pd.concat(
+ [
+ selected_data.loc[special_cols],
+ selected_data.drop(index=special_cols + list(additional_features_dict.keys())).sort_values(
+ sort_by, ascending=order
+ ),
+ selected_data.loc[list(additional_features_dict.keys())].sort_values(sort_by, ascending=order),
+ ]
+ )
return selected_data
@@ -509,26 +511,31 @@ def create_id_card_layout(selected_data: pd.DataFrame, additional_features_dict:
"""
children = []
for _, row in selected_data.iterrows():
- label_style = {
- 'fontWeight': 'bold',
- 'font-style': 'italic'
- } if row["feature_name"] in additional_features_dict.values() else {'fontWeight': 'bold'}
+ label_style = (
+ {"fontWeight": "bold", "font-style": "italic"}
+ if row["feature_name"] in additional_features_dict.values()
+ else {"fontWeight": "bold"}
+ )
children.append(
- dbc.Row([
- dbc.Col(dbc.Label(row["feature_name"]), width=3, style=label_style),
- dbc.Col(dbc.Label(row["feature_value"]), width=5, className="id_card_solid"),
- dbc.Col(width=1),
- dbc.Col(
- dbc.Row(
- dbc.Label(format(row["feature_contrib"], '.4f'), width="auto", style={"padding-top":0}),
- justify="end"
- ),
- width=2,
- className="id_card_solid",
- ) if not np.isnan(row["feature_contrib"]) else None,
- ])
+ dbc.Row(
+ [
+ dbc.Col(dbc.Label(row["feature_name"]), width=3, style=label_style),
+ dbc.Col(dbc.Label(row["feature_value"]), width=5, className="id_card_solid"),
+ dbc.Col(width=1),
+ dbc.Col(
+ dbc.Row(
+ dbc.Label(format(row["feature_contrib"], ".4f"), width="auto", style={"padding-top": 0}),
+ justify="end",
+ ),
+ width=2,
+ className="id_card_solid",
+ )
+ if not np.isnan(row["feature_contrib"])
+ else None,
+ ]
+ )
)
-
+
return children
@@ -550,19 +557,16 @@ def get_feature_filter_options(dataframe: pd.DataFrame, features_dict: dict, spe
Options for the filter
"""
# We use domain name for feature name
- dict_name = [features_dict[i]
- for i in dataframe.drop(special_cols, axis=1).columns]
+ dict_name = [features_dict[i] for i in dataframe.drop(special_cols, axis=1).columns]
dict_id = [i for i in dataframe.drop(special_cols, axis=1).columns]
# Create dataframe to sort it by feature_name
- df_feature_name = pd.DataFrame({'feature_name': dict_name,
- 'feature_id': dict_id})
- df_feature_name = df_feature_name.sort_values(
- by='feature_name').reset_index(drop=True)
+ df_feature_name = pd.DataFrame({"feature_name": dict_name, "feature_id": dict_id})
+ df_feature_name = df_feature_name.sort_values(by="feature_name").reset_index(drop=True)
# Options are sorted by feature_name
- options = [{"label": i, "value": i} for i in special_cols] + \
- [{"label": df_feature_name.loc[i, 'feature_name'],
- "value": df_feature_name.loc[i, 'feature_id']}
- for i in range(len(df_feature_name))]
+ options = [{"label": i, "value": i} for i in special_cols] + [
+ {"label": df_feature_name.loc[i, "feature_name"], "value": df_feature_name.loc[i, "feature_id"]}
+ for i in range(len(df_feature_name))
+ ]
return options
@@ -589,41 +593,36 @@ def create_dropdown_feature_filter(n_clicks_add: Optional[int], options: list) -
index_id = n_clicks_add
# Appending a dropdown block to 'dropdowns_container'children
subset_filter = html.Div(
- id={'type': 'bloc_div',
- 'index': index_id},
+ id={"type": "bloc_div", "index": index_id},
children=[
- html.Div([
- html.Br(),
- # div which will contains label
- html.Div(
- id={'type': 'dynamic-output-label',
- 'index': index_id},
- )
- ]),
- html.Div([
- # div with dopdown button to select feature to filter
- html.Div(dcc.Dropdown(
- id={'type': 'var_dropdown',
- 'index': index_id},
- options=options,
- placeholder="Variable"
- ), style={"width": "30%"}),
- # div which will contains modalities
- html.Div(
- id={'type': 'dynamic-output',
- 'index': index_id},
- style={"width": "50%"}
+ html.Div(
+ [
+ html.Br(),
+ # div which will contains label
+ html.Div(
+ id={"type": "dynamic-output-label", "index": index_id},
),
- # Button to delete bloc
- dbc.Button(
- id={'type': 'del_dropdown_button',
- 'index': index_id},
- children='X',
- color='warning',
- size='sm'
- )
- ], style={'display': 'flex'})
- ]
+ ]
+ ),
+ html.Div(
+ [
+ # div with dopdown button to select feature to filter
+ html.Div(
+ dcc.Dropdown(
+ id={"type": "var_dropdown", "index": index_id}, options=options, placeholder="Variable"
+ ),
+ style={"width": "30%"},
+ ),
+ # div which will contains modalities
+ html.Div(id={"type": "dynamic-output", "index": index_id}, style={"width": "50%"}),
+ # Button to delete bloc
+ dbc.Button(
+ id={"type": "del_dropdown_button", "index": index_id}, children="X", color="warning", size="sm"
+ ),
+ ],
+ style={"display": "flex"},
+ ),
+ ],
)
return subset_filter
@@ -647,64 +646,61 @@ def create_filter_modalities_selection(value: str, id: dict, round_dataframe: pd
Div containing the modalities selection options
"""
if type(round_dataframe[value].iloc[0]) == bool:
- new_element = html.Div(dcc.RadioItems(
- [{'label': val, 'value': val} for
- val in round_dataframe[value].unique()],
- id={'type': 'dynamic-bool',
- 'index': id['index']},
- value=round_dataframe[value].iloc[0],
- inline=False
- ), style={"width": "65%", 'margin-left': '20px'})
- elif (type(round_dataframe[value].iloc[0]) == str) | \
- ((type(round_dataframe[value].iloc[0]) == np.int64) &
- (len(round_dataframe[value].unique()) <= 20)):
- new_element = html.Div(dcc.Dropdown(
- id={
- 'type': 'dynamic-str',
- 'index': id['index']
- },
- options=[{'label': i, 'value': i} for
- i in np.sort(round_dataframe[value].unique())],
- multi=True,
- ), style={"width": "65%", 'margin-left': '20px'})
- elif ((type(round_dataframe[value].iloc[0]) is pd.Timestamp) |
- (type(round_dataframe[value].iloc[0]) is datetime.datetime)):
new_element = html.Div(
- dcc.DatePickerRange(
- id={
- 'type': 'dynamic-date',
- 'index': id['index']
- },
- min_date_allowed=round_dataframe[value].min(),
- max_date_allowed=round_dataframe[value].max(),
- start_date=round_dataframe[value].min(),
- end_date=round_dataframe[value].max()
- ), style={'width': '65%', 'margin-left': '20px'}),
+ dcc.RadioItems(
+ [{"label": val, "value": val} for val in round_dataframe[value].unique()],
+ id={"type": "dynamic-bool", "index": id["index"]},
+ value=round_dataframe[value].iloc[0],
+ inline=False,
+ ),
+ style={"width": "65%", "margin-left": "20px"},
+ )
+ elif (type(round_dataframe[value].iloc[0]) == str) | (
+ (type(round_dataframe[value].iloc[0]) == np.int64) & (len(round_dataframe[value].unique()) <= 20)
+ ):
+ new_element = html.Div(
+ dcc.Dropdown(
+ id={"type": "dynamic-str", "index": id["index"]},
+ options=[{"label": i, "value": i} for i in np.sort(round_dataframe[value].unique())],
+ multi=True,
+ ),
+ style={"width": "65%", "margin-left": "20px"},
+ )
+ elif (type(round_dataframe[value].iloc[0]) is pd.Timestamp) | (
+ type(round_dataframe[value].iloc[0]) is datetime.datetime
+ ):
+ new_element = (
+ html.Div(
+ dcc.DatePickerRange(
+ id={"type": "dynamic-date", "index": id["index"]},
+ min_date_allowed=round_dataframe[value].min(),
+ max_date_allowed=round_dataframe[value].max(),
+ start_date=round_dataframe[value].min(),
+ end_date=round_dataframe[value].max(),
+ ),
+ style={"width": "65%", "margin-left": "20px"},
+ ),
+ )
else:
lower_value = 0
upper_value = 0
- new_element = html.Div([
- dcc.Input(
- id={
- 'type': 'lower',
- 'index': id['index']
- },
- value=lower_value,
- type="number",
- style={'width': '60px'}),
- ' <= {} in [{}, {}]<= '.format(
- value,
- round_dataframe[value].min(),
- round_dataframe[value].max()),
- dcc.Input(
- id={
- 'type': 'upper',
- 'index': id['index']
- },
- value=upper_value,
- type="number",
- style={'width': '60px'}
- )
- ], style={'margin-left': '20px'})
+ new_element = html.Div(
+ [
+ dcc.Input(
+ id={"type": "lower", "index": id["index"]},
+ value=lower_value,
+ type="number",
+ style={"width": "60px"},
+ ),
+ " <= {} in [{}, {}]<= ".format(value, round_dataframe[value].min(), round_dataframe[value].max()),
+ dcc.Input(
+ id={"type": "upper", "index": id["index"]},
+ value=upper_value,
+ type="number",
+ style={"width": "60px"},
+ ),
+ ],
+ style={"margin-left": "20px"},
+ )
return new_element
diff --git a/shapash/webapp/utils/explanations.py b/shapash/webapp/utils/explanations.py
index 1ac70abe..5ab21ee6 100644
--- a/shapash/webapp/utils/explanations.py
+++ b/shapash/webapp/utils/explanations.py
@@ -1,19 +1,18 @@
-# -*- coding: utf-8 -*-
"""
Created on Mon Oct 3 15:12:08 2022
@author: Florine
"""
+
class Explanations:
"""
- Contains the explanations of all "?" buttons in the app.
- ----------
- explanations: object
- SmartApp instance to point to.
+ Contains the explanations of all "?" buttons in the app.
+ ----------
+ explanations: object
+ SmartApp instance to point to.
"""
-
def __init__(self):
self.detail_feature = """
**Local interpretability:** the understanding of the decision
diff --git a/shapash/webapp/utils/utils.py b/shapash/webapp/utils/utils.py
index 5a5726eb..3d84bf48 100644
--- a/shapash/webapp/utils/utils.py
+++ b/shapash/webapp/utils/utils.py
@@ -1,5 +1,5 @@
-import pandas as pd
import numpy as np
+import pandas as pd
def round_to_k(x, k):
@@ -18,7 +18,7 @@ def round_to_k(x, k):
"""
x = float(x)
- new_x = float('%s' % float(f'%.{k}g' % x)) # Rounding to k important figures
+ new_x = float("%s" % float(f"%.{k}g" % x)) # Rounding to k important figures
if new_x % 1 == 0:
return int(new_x) # Avoid the '.0' that can mislead the user that it may be a round number
else:
@@ -60,7 +60,7 @@ def check_row(data, index):
row number corresponding to index
"""
if index is not None:
- df = pd.DataFrame.from_records(data, index='_index_')
+ df = pd.DataFrame.from_records(data, index="_index_")
row = df.index.get_loc(index) if index in list(df.index) else None
else:
row = None
@@ -81,24 +81,26 @@ def split_filter_part(filter_part):
column, operator, value of the filter part
"""
- operators = [['ge ', '>='],
- ['le ', '<='],
- ['lt ', '<'],
- ['gt ', '>'],
- ['ne ', '!='],
- ['eq ', '='],
- ['contains '],
- ['datestartswith ']]
+ operators = [
+ ["ge ", ">="],
+ ["le ", "<="],
+ ["lt ", "<"],
+ ["gt ", ">"],
+ ["ne ", "!="],
+ ["eq ", "="],
+ ["contains "],
+ ["datestartswith "],
+ ]
for operator_type in operators:
for operator in operator_type:
if operator in filter_part:
name_part, value_part = filter_part.split(operator, 1)
- name = name_part[name_part.find('{') + 1: name_part.rfind('}')]
+ name = name_part[name_part.find("{") + 1 : name_part.rfind("}")]
value_part = value_part.strip()
v0 = value_part[0]
- if v0 == value_part[-1] and v0 in ("'", '"', '`'):
- value = value_part[1: -1].replace('\\' + v0, v0)
+ if v0 == value_part[-1] and v0 in ("'", '"', "`"):
+ value = value_part[1:-1].replace("\\" + v0, v0)
else:
try:
value = float(value_part)
@@ -128,17 +130,16 @@ def apply_filter(df, filter_query):
pandas.DataFrame
"""
- filtering_expressions = filter_query.split(' && ')
+ filtering_expressions = filter_query.split(" && ")
for filter_part in filtering_expressions:
col_name, operator, filter_value = split_filter_part(filter_part)
- if operator in ('eq', 'ne', 'lt', 'le', 'gt', 'ge'):
+ if operator in ("eq", "ne", "lt", "le", "gt", "ge"):
# these operators match pandas series operator method names
df = df.loc[getattr(df[col_name], operator)(filter_value)]
- elif operator == 'contains':
+ elif operator == "contains":
df = df.loc[df[col_name].str.contains(filter_value)]
- elif operator == 'datestartswith':
+ elif operator == "datestartswith":
# this is a simplification of the front-end filtering logic,
# only works with complete fields in standard format
df = df.loc[df[col_name].str.startswith(filter_value)]
return df
-
diff --git a/shapash/webapp/webapp_launch.py b/shapash/webapp/webapp_launch.py
index d03cb408..1c768311 100644
--- a/shapash/webapp/webapp_launch.py
+++ b/shapash/webapp/webapp_launch.py
@@ -2,41 +2,42 @@
Webapp launch module
This is an example in python how to launch app from explainer
"""
-from lightgbm import LGBMClassifier, LGBMRegressor
import pandas as pd
+from category_encoders import one_hot
+from lightgbm import LGBMClassifier, LGBMRegressor
from sklearn.model_selection import train_test_split
+
from shapash import SmartExplainer
-from category_encoders import one_hot
cases = {
- 1: 'Titanic regression',
- 2: 'Titanic binary classification',
- 3: 'Titanic multi class classification',
+ 1: "Titanic regression",
+ 2: "Titanic binary classification",
+ 3: "Titanic multi class classification",
}
CASE = 2
-titanic = pd.read_pickle('tests/data/clean_titanic.pkl')
+titanic = pd.read_pickle("tests/data/clean_titanic.pkl")
if CASE == 1:
- features = ['Pclass', 'Survived', 'Embarked', 'Sex']
- encoder = one_hot.OneHotEncoder(titanic, cols=['Embarked', 'Sex'])
+ features = ["Pclass", "Survived", "Embarked", "Sex"]
+ encoder = one_hot.OneHotEncoder(titanic, cols=["Embarked", "Sex"])
X = titanic[features]
- y = titanic['Age'].to_frame()
+ y = titanic["Age"].to_frame()
model = LGBMRegressor()
elif CASE == 2:
- features = ['Pclass', 'Age', 'Embarked', 'Sex']
- encoder = one_hot.OneHotEncoder(titanic, cols=['Embarked', 'Sex'])
+ features = ["Pclass", "Age", "Embarked", "Sex"]
+ encoder = one_hot.OneHotEncoder(titanic, cols=["Embarked", "Sex"])
X = titanic[features]
- y = titanic['Survived'].to_frame()
+ y = titanic["Survived"].to_frame()
model = LGBMClassifier()
else:
- features = ['Survived', 'Age', 'Embarked', 'Sex']
- encoder = one_hot.OneHotEncoder(titanic, cols=['Embarked', 'Sex'])
+ features = ["Survived", "Age", "Embarked", "Sex"]
+ encoder = one_hot.OneHotEncoder(titanic, cols=["Embarked", "Sex"])
X = titanic[features]
- y = titanic['Pclass'].to_frame()
+ y = titanic["Pclass"].to_frame()
model = LGBMClassifier()
titanic_enc = encoder.fit_transform(X)
@@ -55,21 +56,12 @@
model.fit(X_train, y_train)
y_pred = model.predict(X_test)
-y_pred = pd.DataFrame(data=y_pred,
- columns=y.columns.to_list(),
- index=X_test.index)
+y_pred = pd.DataFrame(data=y_pred, columns=y.columns.to_list(), index=X_test.index)
-y_target = pd.DataFrame(data=y_test,
- columns=y.columns.to_list(),
- index=X_test.index)
+y_target = pd.DataFrame(data=y_test, columns=y.columns.to_list(), index=X_test.index)
-response_dict = {0: 'Death', 1: 'Survival'}
-xpl = SmartExplainer(
- model,
- preprocessing=encoder,
- title_story=cases[CASE],
- label_dict=response_dict
-)
+response_dict = {0: "Death", 1: "Survival"}
+xpl = SmartExplainer(model, preprocessing=encoder, title_story=cases[CASE], label_dict=response_dict)
xpl.compile(X_test, y_pred=y_pred, y_target=y_target)
diff --git a/shapash/webapp/webapp_launch_DVF.py b/shapash/webapp/webapp_launch_DVF.py
index 718dfded..c82436d6 100644
--- a/shapash/webapp/webapp_launch_DVF.py
+++ b/shapash/webapp/webapp_launch_DVF.py
@@ -4,39 +4,31 @@
"""
import pandas as pd
+from category_encoders import OrdinalEncoder
from lightgbm import LGBMRegressor
from sklearn.model_selection import train_test_split
-from category_encoders import OrdinalEncoder
-from shapash.data.data_loader import data_loading
from shapash import SmartExplainer
+from shapash.data.data_loader import data_loading
-house_df, house_dict = data_loading('house_prices')
-y_df = house_df['SalePrice'].to_frame()
-X_df = house_df[house_df.columns.difference(['SalePrice'])]
+house_df, house_dict = data_loading("house_prices")
+y_df = house_df["SalePrice"].to_frame()
+X_df = house_df[house_df.columns.difference(["SalePrice"])]
house_df.head()
-categorical_features = [col for col in X_df.columns if X_df[col].dtype == 'object']
-encoder = OrdinalEncoder(
- cols=categorical_features,
- handle_unknown='ignore',
- return_df=True).fit(X_df)
+categorical_features = [col for col in X_df.columns if X_df[col].dtype == "object"]
+encoder = OrdinalEncoder(cols=categorical_features, handle_unknown="ignore", return_df=True).fit(X_df)
X_df = encoder.transform(X_df)
Xtrain, Xtest, ytrain, ytest = train_test_split(X_df, y_df, train_size=0.75, random_state=1)
regressor = LGBMRegressor(n_estimators=200).fit(Xtrain, ytrain)
-y_pred = pd.DataFrame(regressor.predict(Xtest), columns=['pred'], index=Xtest.index)
-y_target = pd.DataFrame(data=ytest,
- columns=y_df.columns.to_list(),
- index=Xtest.index)
+y_pred = pd.DataFrame(regressor.predict(Xtest), columns=["pred"], index=Xtest.index)
+y_target = pd.DataFrame(data=ytest, columns=y_df.columns.to_list(), index=Xtest.index)
xpl = SmartExplainer(
- model=regressor,
- features_dict=house_dict,
- preprocessing=encoder,
- title_story='House Prices - Lightgbm Regressor'
+ model=regressor, features_dict=house_dict, preprocessing=encoder, title_story="House Prices - Lightgbm Regressor"
)
xpl.compile(x=Xtest, y_pred=y_pred, y_target=y_target)
diff --git a/tests/integration_tests/test_contributions_binary.py b/tests/integration_tests/test_contributions_binary.py
index 0c693ddd..f9ce1d52 100644
--- a/tests/integration_tests/test_contributions_binary.py
+++ b/tests/integration_tests/test_contributions_binary.py
@@ -2,10 +2,12 @@
Unit test for contributions binary
"""
import unittest
-from os.path import dirname, abspath, join
-import pandas as pd
-import numpy as np
+from os.path import abspath, dirname, join
+
import category_encoders as ce
+import numpy as np
+import pandas as pd
+
from shapash.decomposition.contributions import inverse_transform_contributions
@@ -18,6 +20,7 @@ class TestContributions(unittest.TestCase):
unittest : [type]
[description]
"""
+
def setUp(self):
"""
Setup
@@ -27,18 +30,18 @@ def setUp(self):
[description]
"""
data_path = dirname(dirname(abspath(__file__)))
- self.ds_titanic_clean = pd.read_pickle(join(data_path, 'data', 'clean_titanic.pkl'))
+ self.ds_titanic_clean = pd.read_pickle(join(data_path, "data", "clean_titanic.pkl"))
def test_inverse_transform_contributions_ce_basen(self):
"""
Unit test inverse transform contributions ce base n
"""
- preprocessing = ce.BaseNEncoder(cols=['Age', 'Sex'], return_df=True, base=3)
+ preprocessing = ce.BaseNEncoder(cols=["Age", "Sex"], return_df=True, base=3)
fitted_dataset = preprocessing.fit_transform(self.ds_titanic_clean)
contributions = pd.DataFrame(
data=np.random.rand(fitted_dataset.shape[0], fitted_dataset.shape[1]),
columns=fitted_dataset.columns,
- index=self.ds_titanic_clean.index
+ index=self.ds_titanic_clean.index,
)
output = inverse_transform_contributions(contributions, preprocessing)
assert isinstance(output, pd.DataFrame)
@@ -49,12 +52,12 @@ def test_inverse_transform_contributions_ce_binary(self):
"""
Unit test inverse transform contributions ce binary
"""
- preprocessing = ce.BinaryEncoder(cols=['Pclass', 'Age', 'Sex'], return_df=True)
+ preprocessing = ce.BinaryEncoder(cols=["Pclass", "Age", "Sex"], return_df=True)
fitted_dataset = preprocessing.fit_transform(self.ds_titanic_clean)
contributions = pd.DataFrame(
data=np.random.rand(fitted_dataset.shape[0], fitted_dataset.shape[1]),
columns=fitted_dataset.columns,
- index=self.ds_titanic_clean.index
+ index=self.ds_titanic_clean.index,
)
output = inverse_transform_contributions(contributions, preprocessing)
assert isinstance(output, pd.DataFrame)
@@ -65,12 +68,12 @@ def test_inverse_transform_contributions_ce_onehot(self):
"""
Unit test inverse transform contributions ce onehot
"""
- preprocessing = ce.OneHotEncoder(cols=['Pclass', 'Sex'], return_df=True)
+ preprocessing = ce.OneHotEncoder(cols=["Pclass", "Sex"], return_df=True)
fitted_dataset = preprocessing.fit_transform(self.ds_titanic_clean)
contributions = pd.DataFrame(
data=np.random.rand(fitted_dataset.shape[0], fitted_dataset.shape[1]),
columns=fitted_dataset.columns,
- index=self.ds_titanic_clean.index
+ index=self.ds_titanic_clean.index,
)
output = inverse_transform_contributions(contributions, preprocessing)
assert isinstance(output, pd.DataFrame)
@@ -81,12 +84,12 @@ def test_inverse_transform_contributions_ce_ordinal(self):
"""
Unit test inverse transform contributions ce ordinal
"""
- preprocessing = ce.OrdinalEncoder(cols=['Pclass', 'Age'], return_df=True)
+ preprocessing = ce.OrdinalEncoder(cols=["Pclass", "Age"], return_df=True)
fitted_dataset = preprocessing.fit_transform(self.ds_titanic_clean)
contributions = pd.DataFrame(
data=np.random.rand(fitted_dataset.shape[0], fitted_dataset.shape[1]),
columns=fitted_dataset.columns,
- index=self.ds_titanic_clean.index
+ index=self.ds_titanic_clean.index,
)
output = inverse_transform_contributions(contributions, preprocessing)
assert isinstance(output, pd.DataFrame)
diff --git a/tests/integration_tests/test_contributions_multiclass.py b/tests/integration_tests/test_contributions_multiclass.py
index 20bd2e8e..0a2e7bc1 100644
--- a/tests/integration_tests/test_contributions_multiclass.py
+++ b/tests/integration_tests/test_contributions_multiclass.py
@@ -2,12 +2,14 @@
Unit test contributions multiclass
"""
import unittest
+
+import numpy as np
+import pandas as pd
import shap
from sklearn.datasets import load_iris
-from sklearn.model_selection import train_test_split
from sklearn.ensemble import RandomForestClassifier
-import numpy as np
-import pandas as pd
+from sklearn.model_selection import train_test_split
+
from shapash.decomposition.contributions import rank_contributions
@@ -24,6 +26,7 @@ class TestContributions(unittest.TestCase):
[type]
[description]
"""
+
def setUp(self):
"""
Setup
@@ -31,11 +34,7 @@ def setUp(self):
iris = load_iris()
x_df = pd.DataFrame(data=iris.data, columns=iris.feature_names)
y_df = pd.DataFrame(data=iris.target, columns=["target"])
- self.x_train, self.x_test, self.y_train, self.y_test = train_test_split(
- x_df,
- y_df,
- random_state=1
- )
+ self.x_train, self.x_test, self.y_train, self.y_test = train_test_split(x_df, y_df, random_state=1)
def check_compute_contributions(self, slist, blist, x_test):
"""
@@ -103,7 +102,4 @@ def test_rank_contributions_1(self):
for i in range(3):
s_ord, x_ord, s_dict = rank_contributions(slist[i], pd.DataFrame(data=self.x_test))
assert np.all(np.diff(np.abs(s_ord), axis=1) <= 0) == 1
- assert np.array_equal(
- x_ord.values,
- np.take_along_axis(self.x_test.values, s_dict.values, axis=1)
- )
+ assert np.array_equal(x_ord.values, np.take_along_axis(self.x_test.values, s_dict.values, axis=1))
diff --git a/tests/integration_tests/test_integration_inverse_tranform.py b/tests/integration_tests/test_integration_inverse_tranform.py
index 75eca8cb..f6a545e5 100644
--- a/tests/integration_tests/test_integration_inverse_tranform.py
+++ b/tests/integration_tests/test_integration_inverse_tranform.py
@@ -2,11 +2,14 @@
Unit test inverse transform
"""
import unittest
-from os.path import dirname, abspath, join
-import pandas as pd
+from os.path import abspath, dirname, join
+
import category_encoders as ce
+import pandas as pd
+
from shapash.utils.transform import inverse_transform
+
class TestInverseTranform(unittest.TestCase):
"""
Unit test invers transform class
@@ -16,6 +19,7 @@ class TestInverseTranform(unittest.TestCase):
unittest : [type]
[description]
"""
+
def setUp(self):
"""
Setup
@@ -25,13 +29,13 @@ def setUp(self):
[description]
"""
data_path = dirname(dirname(abspath(__file__)))
- self.ds_titanic_clean = pd.read_pickle(join(data_path, 'data', 'clean_titanic.pkl'))
+ self.ds_titanic_clean = pd.read_pickle(join(data_path, "data", "clean_titanic.pkl"))
def test_inverse_transform_ce_basen(self):
"""
Unit test inverse transform base n
"""
- preprocessing = ce.BaseNEncoder(cols=['Age', 'Sex'], return_df=True, base=3)
+ preprocessing = ce.BaseNEncoder(cols=["Age", "Sex"], return_df=True, base=3)
fitted_dataset = preprocessing.fit_transform(self.ds_titanic_clean)
output = inverse_transform(fitted_dataset, preprocessing)
pd.testing.assert_frame_equal(output, self.ds_titanic_clean)
@@ -40,7 +44,7 @@ def test_inverse_transform_ce_onehot(self):
"""
Unit test inverse transform ce onehot
"""
- preprocessing = ce.OneHotEncoder(cols=['Age', 'Sex'], return_df=True)
+ preprocessing = ce.OneHotEncoder(cols=["Age", "Sex"], return_df=True)
fitted_dataset = preprocessing.fit_transform(self.ds_titanic_clean)
output = inverse_transform(fitted_dataset, preprocessing)
pd.testing.assert_frame_equal(output, self.ds_titanic_clean)
@@ -49,7 +53,7 @@ def test_inverse_transform_ce_binary(self):
"""
Unit test inverse transform ce binary
"""
- preprocessing = ce.BinaryEncoder(cols=['Age', 'Sex'], return_df=True)
+ preprocessing = ce.BinaryEncoder(cols=["Age", "Sex"], return_df=True)
fitted_dataset = preprocessing.fit_transform(self.ds_titanic_clean)
output = inverse_transform(fitted_dataset, preprocessing)
pd.testing.assert_frame_equal(output, self.ds_titanic_clean)
@@ -58,7 +62,7 @@ def test_inverse_transform_ce_ordinal(self):
"""
Unit test inverse transform ce ordinal
"""
- preprocessing = ce.OrdinalEncoder(cols=['Age', 'Sex'], return_df=True)
+ preprocessing = ce.OrdinalEncoder(cols=["Age", "Sex"], return_df=True)
fitted_dataset = preprocessing.fit_transform(self.ds_titanic_clean)
output = inverse_transform(fitted_dataset, preprocessing)
pd.testing.assert_frame_equal(output, self.ds_titanic_clean)
diff --git a/tests/integration_tests/test_report_generation.py b/tests/integration_tests/test_report_generation.py
index 4bfbedd9..91a987f7 100644
--- a/tests/integration_tests/test_report_generation.py
+++ b/tests/integration_tests/test_report_generation.py
@@ -1,35 +1,35 @@
-import unittest
-import tempfile
-import shutil
-import numpy as np
import os
-import pandas as pd
+import shutil
+import tempfile
+import unittest
+
import catboost as cb
import category_encoders as ce
+import numpy as np
+import pandas as pd
+from category_encoders import OrdinalEncoder
from shapash import SmartExplainer
from shapash.report.generation import execute_report, export_and_save_report
-from category_encoders import OrdinalEncoder
current_path = os.path.dirname(os.path.abspath(__file__))
class TestGeneration(unittest.TestCase):
-
def setUp(self):
- df = pd.DataFrame(range(0, 21), columns=['id'])
- df['y'] = df['id'].apply(lambda x: 1 if x < 10 else 0)
- df['x1'] = np.random.randint(1, 123, df.shape[0])
- df['x2'] = np.random.randint(1, 3, df.shape[0])
- df['x3'] = np.random.choice(['A', 'B', 'C', 'D'], df.shape[0])
- df['x4'] = np.random.choice(['A', 'B', 'C', np.nan], df.shape[0])
- df = df.set_index('id')
+ df = pd.DataFrame(range(0, 21), columns=["id"])
+ df["y"] = df["id"].apply(lambda x: 1 if x < 10 else 0)
+ df["x1"] = np.random.randint(1, 123, df.shape[0])
+ df["x2"] = np.random.randint(1, 3, df.shape[0])
+ df["x3"] = np.random.choice(["A", "B", "C", "D"], df.shape[0])
+ df["x4"] = np.random.choice(["A", "B", "C", np.nan], df.shape[0])
+ df = df.set_index("id")
encoder = ce.OrdinalEncoder(cols=["x3", "x4"], handle_unknown="None")
encoder_fitted = encoder.fit(df)
df_encoded = encoder_fitted.transform(df)
- clf = cb.CatBoostClassifier(n_estimators=1).fit(df_encoded[['x1', 'x2', 'x3', 'x4']], df_encoded['y'])
+ clf = cb.CatBoostClassifier(n_estimators=1).fit(df_encoded[["x1", "x2", "x3", "x4"]], df_encoded["y"])
self.xpl = SmartExplainer(model=clf, preprocessing=encoder)
- self.xpl.compile(x=df_encoded[['x1', 'x2', 'x3', 'x4']])
+ self.xpl.compile(x=df_encoded[["x1", "x2", "x3", "x4"]])
self.df = df_encoded
def test_execute_report_1(self):
@@ -38,12 +38,12 @@ def test_execute_report_1(self):
execute_report(
working_dir=tmp_dir_path,
explainer=self.xpl,
- project_info_file=os.path.join(current_path, '../data/metadata.yaml'),
+ project_info_file=os.path.join(current_path, "../data/metadata.yaml"),
config=None,
- notebook_path=None
+ notebook_path=None,
)
- assert os.path.exists(os.path.join(tmp_dir_path, 'smart_explainer.pickle'))
- assert os.path.exists(os.path.join(tmp_dir_path, 'base_report.ipynb'))
+ assert os.path.exists(os.path.join(tmp_dir_path, "smart_explainer.pickle"))
+ assert os.path.exists(os.path.join(tmp_dir_path, "base_report.ipynb"))
shutil.rmtree(tmp_dir_path)
@@ -53,14 +53,14 @@ def test_execute_report_2(self):
execute_report(
working_dir=tmp_dir_path,
explainer=self.xpl,
- project_info_file=os.path.join(current_path, '../data/metadata.yaml'),
- x_train=self.df[['x1', 'x2', 'x3', 'x4']],
+ project_info_file=os.path.join(current_path, "../data/metadata.yaml"),
+ x_train=self.df[["x1", "x2", "x3", "x4"]],
config=None,
- notebook_path=None
+ notebook_path=None,
)
- assert os.path.exists(os.path.join(tmp_dir_path, 'x_train.csv'))
- assert os.path.exists(os.path.join(tmp_dir_path, 'smart_explainer.pickle'))
- assert os.path.exists(os.path.join(tmp_dir_path, 'base_report.ipynb'))
+ assert os.path.exists(os.path.join(tmp_dir_path, "x_train.csv"))
+ assert os.path.exists(os.path.join(tmp_dir_path, "smart_explainer.pickle"))
+ assert os.path.exists(os.path.join(tmp_dir_path, "base_report.ipynb"))
shutil.rmtree(tmp_dir_path)
@@ -70,16 +70,16 @@ def test_execute_report_3(self):
execute_report(
working_dir=tmp_dir_path,
explainer=self.xpl,
- project_info_file=os.path.join(current_path, '../data/metadata.yaml'),
- x_train=self.df[['x1', 'x2', 'x3', 'x4']],
- y_test=self.df['y'],
+ project_info_file=os.path.join(current_path, "../data/metadata.yaml"),
+ x_train=self.df[["x1", "x2", "x3", "x4"]],
+ y_test=self.df["y"],
config=None,
- notebook_path=None
+ notebook_path=None,
)
- assert os.path.exists(os.path.join(tmp_dir_path, 'x_train.csv'))
- assert os.path.exists(os.path.join(tmp_dir_path, 'y_test.csv'))
- assert os.path.exists(os.path.join(tmp_dir_path, 'smart_explainer.pickle'))
- assert os.path.exists(os.path.join(tmp_dir_path, 'base_report.ipynb'))
+ assert os.path.exists(os.path.join(tmp_dir_path, "x_train.csv"))
+ assert os.path.exists(os.path.join(tmp_dir_path, "y_test.csv"))
+ assert os.path.exists(os.path.join(tmp_dir_path, "smart_explainer.pickle"))
+ assert os.path.exists(os.path.join(tmp_dir_path, "base_report.ipynb"))
shutil.rmtree(tmp_dir_path)
@@ -89,18 +89,18 @@ def test_execute_report_4(self):
execute_report(
working_dir=tmp_dir_path,
explainer=self.xpl,
- project_info_file=os.path.join(current_path, '../data/metadata.yaml'),
- x_train=self.df[['x1', 'x2', 'x3', 'x4']],
- y_train=self.df['y'],
- y_test=self.df['y'],
+ project_info_file=os.path.join(current_path, "../data/metadata.yaml"),
+ x_train=self.df[["x1", "x2", "x3", "x4"]],
+ y_train=self.df["y"],
+ y_test=self.df["y"],
config=None,
- notebook_path=None
+ notebook_path=None,
)
- assert os.path.exists(os.path.join(tmp_dir_path, 'x_train.csv'))
- assert os.path.exists(os.path.join(tmp_dir_path, 'y_test.csv'))
- assert os.path.exists(os.path.join(tmp_dir_path, 'y_train.csv'))
- assert os.path.exists(os.path.join(tmp_dir_path, 'smart_explainer.pickle'))
- assert os.path.exists(os.path.join(tmp_dir_path, 'base_report.ipynb'))
+ assert os.path.exists(os.path.join(tmp_dir_path, "x_train.csv"))
+ assert os.path.exists(os.path.join(tmp_dir_path, "y_test.csv"))
+ assert os.path.exists(os.path.join(tmp_dir_path, "y_train.csv"))
+ assert os.path.exists(os.path.join(tmp_dir_path, "smart_explainer.pickle"))
+ assert os.path.exists(os.path.join(tmp_dir_path, "base_report.ipynb"))
shutil.rmtree(tmp_dir_path)
@@ -111,18 +111,18 @@ def test_execute_report_5(self):
execute_report(
working_dir=tmp_dir_path,
explainer=self.xpl,
- project_info_file=os.path.join(current_path, '../data/metadata.yaml'),
- x_train=self.df[['x1', 'x2', 'x3', 'x4']],
- y_train=self.df['y'],
- y_test=self.df['y'],
- notebook_path=None
+ project_info_file=os.path.join(current_path, "../data/metadata.yaml"),
+ x_train=self.df[["x1", "x2", "x3", "x4"]],
+ y_train=self.df["y"],
+ y_test=self.df["y"],
+ notebook_path=None,
)
self.xpl.palette_name = "default"
- assert os.path.exists(os.path.join(tmp_dir_path, 'x_train.csv'))
- assert os.path.exists(os.path.join(tmp_dir_path, 'y_test.csv'))
- assert os.path.exists(os.path.join(tmp_dir_path, 'y_train.csv'))
- assert os.path.exists(os.path.join(tmp_dir_path, 'smart_explainer.pickle'))
- assert os.path.exists(os.path.join(tmp_dir_path, 'base_report.ipynb'))
+ assert os.path.exists(os.path.join(tmp_dir_path, "x_train.csv"))
+ assert os.path.exists(os.path.join(tmp_dir_path, "y_test.csv"))
+ assert os.path.exists(os.path.join(tmp_dir_path, "y_train.csv"))
+ assert os.path.exists(os.path.join(tmp_dir_path, "smart_explainer.pickle"))
+ assert os.path.exists(os.path.join(tmp_dir_path, "base_report.ipynb"))
shutil.rmtree(tmp_dir_path)
@@ -132,10 +132,10 @@ def test_export_and_save_report_1(self):
execute_report(
working_dir=tmp_dir_path,
explainer=self.xpl,
- project_info_file=os.path.join(current_path, '../data/metadata.yaml'),
+ project_info_file=os.path.join(current_path, "../data/metadata.yaml"),
)
- outfile = os.path.join(tmp_dir_path, 'report.html')
+ outfile = os.path.join(tmp_dir_path, "report.html")
export_and_save_report(working_dir=tmp_dir_path, output_file=outfile)
assert os.path.exists(outfile)
shutil.rmtree(tmp_dir_path)
diff --git a/tests/unit_tests/backend/test_base_backend.py b/tests/unit_tests/backend/test_base_backend.py
index 818858e4..1c22bb74 100644
--- a/tests/unit_tests/backend/test_base_backend.py
+++ b/tests/unit_tests/backend/test_base_backend.py
@@ -1,20 +1,20 @@
+import types
import unittest
from unittest.mock import patch
-import types
+import category_encoders as ce
import numpy as np
import pandas as pd
-import category_encoders as ce
-from sklearn.ensemble import RandomForestRegressor, RandomForestClassifier
from pandas.testing import assert_frame_equal, assert_series_equal
+from sklearn.ensemble import RandomForestClassifier, RandomForestRegressor
from shapash.backend.base_backend import BaseBackend, _needs_preprocessing
from shapash.explainer.smart_state import SmartState
class TestBackend(BaseBackend):
- column_aggregation = 'sum'
- name = 'test'
+ column_aggregation = "sum"
+ name = "test"
def run_explainer(self, x):
return [[0, 1]]
@@ -22,15 +22,13 @@ def run_explainer(self, x):
class TestBaseBackend(unittest.TestCase):
def predict(self, arg1, arg2):
- matrx = np.array(
- [12, 3, 7]
- )
+ matrx = np.array([12, 3, 7])
return matrx
def setUp(self):
self.model = lambda: None
self.model.predict = types.MethodType(self.predict, self.model)
- self.preprocessing = ce.OneHotEncoder(cols=['Onehot1', 'Onehot2'])
+ self.preprocessing = ce.OneHotEncoder(cols=["Onehot1", "Onehot2"])
self.test_backend = TestBackend(model=self.model, preprocessing=self.preprocessing)
@@ -45,7 +43,7 @@ def test_init(self):
def test_run_explainer(self):
assert self.test_backend.run_explainer(pd.DataFrame([0])) == [[0, 1]]
- @patch('shapash.backend.base_backend.BaseBackend.format_and_aggregate_local_contributions')
+ @patch("shapash.backend.base_backend.BaseBackend.format_and_aggregate_local_contributions")
def test_get_local_contributions(self, mock_format_contrib):
mock_format_contrib.return_value = [0, 4]
res = self.test_backend.get_local_contributions(pd.DataFrame([0]), dict(contributions=[np.array([0])]))
@@ -67,51 +65,47 @@ def test_get_global_features_importance(self):
assert_series_equal(res, pd.Series([0.1, 0.2, 0.3, 0.4]))
def test_format_and_aggregate_local_contributions(self):
- self.test_backend._case = 'classification'
+ self.test_backend._case = "classification"
self.test_backend._classes = [0, 1]
self.test_backend.preprocessing = None
- x = pd.DataFrame([[0, 1, 2], [3, 4, 5]], columns=['a', 'b', 'c'])
+ x = pd.DataFrame([[0, 1, 2], [3, 4, 5]], columns=["a", "b", "c"])
contributions = [np.array([[-0.1, 0.2, -0.3], [0.1, -0.2, 0.3]]), np.array([[0.1, 0.2, 0.3], [0.1, 0.2, 0.3]])]
res_contributions = self.test_backend.format_and_aggregate_local_contributions(x, contributions)
assert len(res_contributions) == 2
assert_frame_equal(
- res_contributions[0],
- pd.DataFrame([[-0.1, 0.2, -0.3], [0.1, -0.2, 0.3]], columns=['a', 'b', 'c'])
+ res_contributions[0], pd.DataFrame([[-0.1, 0.2, -0.3], [0.1, -0.2, 0.3]], columns=["a", "b", "c"])
)
assert_frame_equal(
- res_contributions[1],
- pd.DataFrame([[0.1, 0.2, 0.3], [0.1, 0.2, 0.3]], columns=['a', 'b', 'c'])
+ res_contributions[1], pd.DataFrame([[0.1, 0.2, 0.3], [0.1, 0.2, 0.3]], columns=["a", "b", "c"])
)
def test_format_and_aggregate_local_contributions_2(self):
- self.test_backend._case = 'classification'
+ self.test_backend._case = "classification"
self.test_backend._classes = [0, 1]
self.test_backend.preprocessing = None
- x = pd.DataFrame([[0, 1, 2], [3, 4, 5]], columns=['a', 'b', 'c'])
+ x = pd.DataFrame([[0, 1, 2], [3, 4, 5]], columns=["a", "b", "c"])
contributions = [
- pd.DataFrame([[-0.1, 0.2, -0.3], [0.1, -0.2, 0.3]], columns=['a', 'b', 'c']),
- pd.DataFrame([[0.1, 0.2, 0.3], [0.1, 0.2, 0.3]], columns=['a', 'b', 'c'])
+ pd.DataFrame([[-0.1, 0.2, -0.3], [0.1, -0.2, 0.3]], columns=["a", "b", "c"]),
+ pd.DataFrame([[0.1, 0.2, 0.3], [0.1, 0.2, 0.3]], columns=["a", "b", "c"]),
]
res_contributions = self.test_backend.format_and_aggregate_local_contributions(x, contributions)
assert len(res_contributions) == 2
assert_frame_equal(
- res_contributions[0],
- pd.DataFrame([[-0.1, 0.2, -0.3], [0.1, -0.2, 0.3]], columns=['a', 'b', 'c'])
+ res_contributions[0], pd.DataFrame([[-0.1, 0.2, -0.3], [0.1, -0.2, 0.3]], columns=["a", "b", "c"])
)
assert_frame_equal(
- res_contributions[1],
- pd.DataFrame([[0.1, 0.2, 0.3], [0.1, 0.2, 0.3]], columns=['a', 'b', 'c'])
+ res_contributions[1], pd.DataFrame([[0.1, 0.2, 0.3], [0.1, 0.2, 0.3]], columns=["a", "b", "c"])
)
def test_needs_preprocessing(self):
- df = pd.DataFrame({'Onehot1': ['A', 'B', 'A', 'B'], 'Onehot2': ['C', 'D', 'C', 'D']})
- encoder = ce.OneHotEncoder(cols=['Onehot1', 'Onehot2']).fit(df)
+ df = pd.DataFrame({"Onehot1": ["A", "B", "A", "B"], "Onehot2": ["C", "D", "C", "D"]})
+ encoder = ce.OneHotEncoder(cols=["Onehot1", "Onehot2"]).fit(df)
df_ohe = encoder.transform(df)
res = _needs_preprocessing(df_ohe.columns, df_ohe, preprocessing=encoder)
@@ -119,8 +113,8 @@ def test_needs_preprocessing(self):
assert res is True
def test_needs_preprocessing_2(self):
- df = pd.DataFrame({'Onehot1': ['A', 'B', 'A', 'B'], 'Onehot2': ['C', 'D', 'C', 'D']})
- encoder = ce.OrdinalEncoder(cols=['Onehot1', 'Onehot2']).fit(df)
+ df = pd.DataFrame({"Onehot1": ["A", "B", "A", "B"], "Onehot2": ["C", "D", "C", "D"]})
+ encoder = ce.OrdinalEncoder(cols=["Onehot1", "Onehot2"]).fit(df)
df_ord = encoder.transform(df)
res = _needs_preprocessing(df_ord.columns, df_ord, preprocessing=encoder)
diff --git a/tests/unit_tests/backend/test_lime_backend.py b/tests/unit_tests/backend/test_lime_backend.py
index 9d4a719b..887eaa28 100644
--- a/tests/unit_tests/backend/test_lime_backend.py
+++ b/tests/unit_tests/backend/test_lime_backend.py
@@ -3,11 +3,13 @@
"""
import unittest
+
+import category_encoders as ce
import numpy as np
import pandas as pd
import sklearn.ensemble as ske
import xgboost as xgb
-import category_encoders as ce
+
from shapash.backend.lime_backend import LimeBackend
@@ -17,29 +19,29 @@ def setUp(self):
xgb.XGBClassifier(n_estimators=1),
ske.RandomForestClassifier(n_estimators=1),
ske.RandomForestRegressor(n_estimators=1),
- ske.GradientBoostingRegressor(n_estimators=1)
+ ske.GradientBoostingRegressor(n_estimators=1),
]
- df = pd.DataFrame(range(0, 4), columns=['id'])
- df['y'] = df['id'].apply(lambda x: 1 if x < 3 else 0)
- df['x1'] = np.random.randint(1, 123, df.shape[0])
- df['x2'] = np.random.randint(1, 3, df.shape[0])
- df = df.set_index('id')
- self.x_df = df[['x1', 'x2']]
- self.y_df = df['y'].to_frame()
+ df = pd.DataFrame(range(0, 4), columns=["id"])
+ df["y"] = df["id"].apply(lambda x: 1 if x < 3 else 0)
+ df["x1"] = np.random.randint(1, 123, df.shape[0])
+ df["x2"] = np.random.randint(1, 3, df.shape[0])
+ df = df.set_index("id")
+ self.x_df = df[["x1", "x2"]]
+ self.y_df = df["y"].to_frame()
def test_init(self):
for model in self.model_list:
print(type(model))
model.fit(self.x_df, self.y_df)
backend_xpl = LimeBackend(model)
- assert hasattr(backend_xpl, 'explainer')
+ assert hasattr(backend_xpl, "explainer")
backend_xpl = LimeBackend(model, data=self.x_df)
- assert hasattr(backend_xpl, 'data')
+ assert hasattr(backend_xpl, "data")
backend_xpl = LimeBackend(model, preprocessing=ce.OrdinalEncoder())
- assert hasattr(backend_xpl, 'preprocessing')
+ assert hasattr(backend_xpl, "preprocessing")
assert isinstance(backend_xpl.preprocessing, ce.OrdinalEncoder)
def test_get_global_contributions(self):
diff --git a/tests/unit_tests/backend/test_shap_backend.py b/tests/unit_tests/backend/test_shap_backend.py
index fe87e265..49790cd7 100644
--- a/tests/unit_tests/backend/test_shap_backend.py
+++ b/tests/unit_tests/backend/test_shap_backend.py
@@ -1,10 +1,11 @@
import unittest
+
+import catboost as cb
+import lightgbm as lgb
import numpy as np
import pandas as pd
import sklearn.ensemble as ske
import xgboost as xgb
-import lightgbm as lgb
-import catboost as cb
from shapash.backend.shap_backend import ShapBackend
@@ -12,21 +13,27 @@
class TestShapBackend(unittest.TestCase):
def setUp(self):
self.model_list = [
- lgb.LGBMRegressor(n_estimators=1), lgb.LGBMClassifier(n_estimators=1),
- xgb.XGBRegressor(n_estimators=1), xgb.XGBClassifier(n_estimators=1),
- cb.CatBoostRegressor(n_estimators=1), cb.CatBoostClassifier(n_estimators=1),
- ske.GradientBoostingRegressor(n_estimators=1), ske.GradientBoostingClassifier(n_estimators=1),
- ske.ExtraTreesRegressor(n_estimators=1), ske.ExtraTreesClassifier(n_estimators=1),
- ske.RandomForestRegressor(n_estimators=1), ske.RandomForestClassifier(n_estimators=1)
+ lgb.LGBMRegressor(n_estimators=1),
+ lgb.LGBMClassifier(n_estimators=1),
+ xgb.XGBRegressor(n_estimators=1),
+ xgb.XGBClassifier(n_estimators=1),
+ cb.CatBoostRegressor(n_estimators=1),
+ cb.CatBoostClassifier(n_estimators=1),
+ ske.GradientBoostingRegressor(n_estimators=1),
+ ske.GradientBoostingClassifier(n_estimators=1),
+ ske.ExtraTreesRegressor(n_estimators=1),
+ ske.ExtraTreesClassifier(n_estimators=1),
+ ske.RandomForestRegressor(n_estimators=1),
+ ske.RandomForestClassifier(n_estimators=1),
]
- df = pd.DataFrame(range(0, 21), columns=['id'])
- df['y'] = df['id'].apply(lambda x: 1 if x < 10 else 0)
- df['x1'] = np.random.randint(1, 123, df.shape[0])
- df['x2'] = np.random.randint(1, 3, df.shape[0])
- df = df.set_index('id')
- self.x_df = df[['x1', 'x2']]
- self.y_df = df['y'].to_frame()
+ df = pd.DataFrame(range(0, 21), columns=["id"])
+ df["y"] = df["id"].apply(lambda x: 1 if x < 10 else 0)
+ df["x1"] = np.random.randint(1, 123, df.shape[0])
+ df["x2"] = np.random.randint(1, 3, df.shape[0])
+ df = df.set_index("id")
+ self.x_df = df[["x1", "x2"]]
+ self.y_df = df["y"].to_frame()
def test_shap_backend_init(self):
"""
@@ -36,7 +43,7 @@ def test_shap_backend_init(self):
print(type(model))
model.fit(self.x_df, self.y_df)
backend_xpl = ShapBackend(model)
- assert hasattr(backend_xpl, 'explainer')
+ assert hasattr(backend_xpl, "explainer")
def test_run_explainer(self):
for model in self.model_list:
diff --git a/tests/unit_tests/data/test_data_loader.py b/tests/unit_tests/data/test_data_loader.py
index ceaaf9ec..2a9d19fb 100644
--- a/tests/unit_tests/data/test_data_loader.py
+++ b/tests/unit_tests/data/test_data_loader.py
@@ -2,17 +2,18 @@
Unit test data loader
"""
import unittest
-from shapash.data.data_loader import data_loading
+
import pandas as pd
+from shapash.data.data_loader import data_loading
-class Test_load_data(unittest.TestCase):
+class Test_load_data(unittest.TestCase):
def test_load_house(self):
"""
Unit test house_prices
"""
- house_df, house_dict = data_loading('house_prices')
+ house_df, house_dict = data_loading("house_prices")
assert isinstance(house_df, pd.DataFrame)
assert isinstance(house_dict, dict)
@@ -20,7 +21,7 @@ def test_load_titanic(self):
"""
Unit test load_titanic
"""
- titanic_df, titanic_dict = data_loading('titanic')
+ titanic_df, titanic_dict = data_loading("titanic")
assert isinstance(titanic_df, pd.DataFrame)
assert isinstance(titanic_dict, dict)
@@ -28,12 +29,12 @@ def test_load_telco(self):
"""
Unit test telco_customer
"""
- telco_df = data_loading('telco_customer_churn')
+ telco_df = data_loading("telco_customer_churn")
assert isinstance(telco_df, pd.DataFrame)
def test_load_accidents(self):
"""
Unit test us_car_accident
"""
- accidents_df = data_loading('us_car_accident')
+ accidents_df = data_loading("us_car_accident")
assert isinstance(accidents_df, pd.DataFrame)
diff --git a/tests/unit_tests/decomposition/test_contributions.py b/tests/unit_tests/decomposition/test_contributions.py
index 65aafe79..12ae3957 100644
--- a/tests/unit_tests/decomposition/test_contributions.py
+++ b/tests/unit_tests/decomposition/test_contributions.py
@@ -2,11 +2,13 @@
Unit test of contributions
"""
import unittest
-import pandas as pd
+
import numpy as np
+import pandas as pd
+
from shapash.decomposition.contributions import rank_contributions
-
+
class TestContributions(unittest.TestCase):
"""
Unit test class of contributions
@@ -22,38 +24,29 @@ def test_rank_contributions_1(self):
Unit test rank contributions 1
"""
dataframe_s = pd.DataFrame(
- [[3.4, 1, -9, 4],
- [-45, 3, 43, -9]],
- columns=["Phi_" + str(i) for i in range(4)],
- index=['raw_1', 'raw_2']
+ [[3.4, 1, -9, 4], [-45, 3, 43, -9]], columns=["Phi_" + str(i) for i in range(4)], index=["raw_1", "raw_2"]
)
dataframe_x = pd.DataFrame(
- [['Male', 'House', 'Married', 'PhD'],
- ['Female', 'Flat', 'Married', 'Master']],
+ [["Male", "House", "Married", "PhD"], ["Female", "Flat", "Married", "Master"]],
columns=["X" + str(i) for i in range(4)],
- index=['raw_1', 'raw_2']
+ index=["raw_1", "raw_2"],
)
expected_s_ord = pd.DataFrame(
- data=[[-9, 4, 3.4, 1],
- [-45, 43, -9, 3]],
- columns=['contribution_' + str(i) for i in range(4)],
- index=['raw_1', 'raw_2']
+ data=[[-9, 4, 3.4, 1], [-45, 43, -9, 3]],
+ columns=["contribution_" + str(i) for i in range(4)],
+ index=["raw_1", "raw_2"],
)
expected_x_ord = pd.DataFrame(
- data=[['Married', 'PhD', 'Male', 'House'],
- ['Female', 'Married', 'Master', 'Flat']],
- columns=['feature_' + str(i) for i in range(4)],
- index=['raw_1', 'raw_2']
+ data=[["Married", "PhD", "Male", "House"], ["Female", "Married", "Master", "Flat"]],
+ columns=["feature_" + str(i) for i in range(4)],
+ index=["raw_1", "raw_2"],
)
expected_s_dict = pd.DataFrame(
- data=[[2, 3, 0, 1],
- [0, 2, 3, 1]],
- columns=['feature_' + str(i) for i in range(4)],
- index=['raw_1', 'raw_2']
+ data=[[2, 3, 0, 1], [0, 2, 3, 1]], columns=["feature_" + str(i) for i in range(4)], index=["raw_1", "raw_2"]
)
s_ord, x_ord, s_dict = rank_contributions(dataframe_s, dataframe_x)
@@ -68,4 +61,4 @@ def test_rank_contributions_1(self):
assert pd.Index.equals(s_ord.index, expected_s_ord.index)
assert pd.Index.equals(x_ord.index, expected_x_ord.index)
- assert pd.Index.equals(s_dict.index, expected_s_dict.index)
\ No newline at end of file
+ assert pd.Index.equals(s_dict.index, expected_s_dict.index)
diff --git a/tests/unit_tests/decomposition/test_inverse_contribution.py b/tests/unit_tests/decomposition/test_inverse_contribution.py
index 4ab358a1..dea484d3 100644
--- a/tests/unit_tests/decomposition/test_inverse_contribution.py
+++ b/tests/unit_tests/decomposition/test_inverse_contribution.py
@@ -2,11 +2,13 @@
Unit test of Inverse Transform
"""
import unittest
-import pandas as pd
-import numpy as np
+
import category_encoders as ce
+import numpy as np
+import pandas as pd
import sklearn.preprocessing as skp
from sklearn.compose import ColumnTransformer
+
from shapash.decomposition.contributions import inverse_transform_contributions
@@ -24,51 +26,73 @@ def test_multiple_encoding_category_encoder(self):
"""
Test multiple preprocessing
"""
- train = pd.DataFrame({'Onehot1': ['A', 'B'], 'Onehot2': ['C', 'D'],
- 'Binary1': ['E', 'F'], 'Binary2': ['G', 'H'],
- 'Ordinal1': ['I', 'J'], 'Ordinal2': ['K', 'L'],
- 'BaseN1': ['M', 'N'], 'BaseN2': ['O', 'P'],
- 'Target1': ['Q', 'R'], 'Target2': ['S', 'T'],
- 'other': ['other', np.nan]})
-
- contributions = pd.DataFrame([[1, 0, 1, 1, 3, 0, -3.5, 0, 4, 5, 0, 6, 7, 0, 8, 9, 10],
- [.5, .5, 2, 0, 1.5, 1.5, 5.5, -2, -4, -5, 8.5, -2.5, -7, 14, -8, -9, -10]],
- index=['index1', 'index2'])
-
- expected_contrib = pd.DataFrame({'Onehot1': [1., 1.], 'Onehot2': [2, 2],
- 'Binary1': [3., 3.], 'Binary2': [-3.5, 3.5],
- 'Ordinal1': [4, -4], 'Ordinal2': [5, -5],
- 'BaseN1': [6., 6.], 'BaseN2': [7, 7],
- 'Target1': [8, -8], 'Target2': [9, -9],
- 'other': [10, -10]},
- index=['index1', 'index2'])
-
- y = pd.DataFrame(data=[0, 1], columns=['y'])
-
- enc_onehot = ce.OneHotEncoder(cols=['Onehot1', 'Onehot2']).fit(train)
+ train = pd.DataFrame(
+ {
+ "Onehot1": ["A", "B"],
+ "Onehot2": ["C", "D"],
+ "Binary1": ["E", "F"],
+ "Binary2": ["G", "H"],
+ "Ordinal1": ["I", "J"],
+ "Ordinal2": ["K", "L"],
+ "BaseN1": ["M", "N"],
+ "BaseN2": ["O", "P"],
+ "Target1": ["Q", "R"],
+ "Target2": ["S", "T"],
+ "other": ["other", np.nan],
+ }
+ )
+
+ contributions = pd.DataFrame(
+ [
+ [1, 0, 1, 1, 3, 0, -3.5, 0, 4, 5, 0, 6, 7, 0, 8, 9, 10],
+ [0.5, 0.5, 2, 0, 1.5, 1.5, 5.5, -2, -4, -5, 8.5, -2.5, -7, 14, -8, -9, -10],
+ ],
+ index=["index1", "index2"],
+ )
+
+ expected_contrib = pd.DataFrame(
+ {
+ "Onehot1": [1.0, 1.0],
+ "Onehot2": [2, 2],
+ "Binary1": [3.0, 3.0],
+ "Binary2": [-3.5, 3.5],
+ "Ordinal1": [4, -4],
+ "Ordinal2": [5, -5],
+ "BaseN1": [6.0, 6.0],
+ "BaseN2": [7, 7],
+ "Target1": [8, -8],
+ "Target2": [9, -9],
+ "other": [10, -10],
+ },
+ index=["index1", "index2"],
+ )
+
+ y = pd.DataFrame(data=[0, 1], columns=["y"])
+
+ enc_onehot = ce.OneHotEncoder(cols=["Onehot1", "Onehot2"]).fit(train)
train_onehot = enc_onehot.transform(train)
- enc_binary = ce.BinaryEncoder(cols=['Binary1', 'Binary2']).fit(train_onehot)
+ enc_binary = ce.BinaryEncoder(cols=["Binary1", "Binary2"]).fit(train_onehot)
train_binary = enc_binary.transform(train_onehot)
- enc_ordinal = ce.OrdinalEncoder(cols=['Ordinal1', 'Ordinal2']).fit(train_binary)
+ enc_ordinal = ce.OrdinalEncoder(cols=["Ordinal1", "Ordinal2"]).fit(train_binary)
train_ordinal = enc_ordinal.transform(train_binary)
- enc_basen = ce.BaseNEncoder(cols=['BaseN1', 'BaseN2']).fit(train_ordinal)
+ enc_basen = ce.BaseNEncoder(cols=["BaseN1", "BaseN2"]).fit(train_ordinal)
train_basen = enc_basen.transform(train_ordinal)
- enc_target = ce.TargetEncoder(cols=['Target1', 'Target2']).fit(train_basen, y)
+ enc_target = ce.TargetEncoder(cols=["Target1", "Target2"]).fit(train_basen, y)
input_dict1 = dict()
- input_dict1['col'] = 'Onehot2'
- input_dict1['mapping'] = pd.Series(data=['C', 'D', np.nan], index=['C', 'D', 'missing'])
- input_dict1['data_type'] = 'object'
+ input_dict1["col"] = "Onehot2"
+ input_dict1["mapping"] = pd.Series(data=["C", "D", np.nan], index=["C", "D", "missing"])
+ input_dict1["data_type"] = "object"
input_dict2 = dict()
- input_dict2['col'] = 'Binary2'
- input_dict2['mapping'] = pd.Series(data=['G', 'H', np.nan], index=['G', 'H', 'missing'])
- input_dict2['data_type'] = 'object'
+ input_dict2["col"] = "Binary2"
+ input_dict2["mapping"] = pd.Series(data=["G", "H", np.nan], index=["G", "H", "missing"])
+ input_dict2["data_type"] = "object"
input_dict3 = dict()
- input_dict3['col'] = 'Ordinal2'
- input_dict3['mapping'] = pd.Series(data=['K', 'L', np.nan], index=['K', 'L', 'missing'])
- input_dict3['data_type'] = 'object'
+ input_dict3["col"] = "Ordinal2"
+ input_dict3["mapping"] = pd.Series(data=["K", "L", np.nan], index=["K", "L", "missing"])
+ input_dict3["data_type"] = "object"
list_dict = [input_dict2, input_dict3]
result1 = enc_onehot.transform(train)
@@ -79,10 +103,9 @@ def test_multiple_encoding_category_encoder(self):
contributions.columns = result5.columns
- original = inverse_transform_contributions(contributions,
- [enc_onehot, enc_binary, enc_ordinal, enc_basen, enc_target,
- input_dict1,
- list_dict])
+ original = inverse_transform_contributions(
+ contributions, [enc_onehot, enc_binary, enc_ordinal, enc_basen, enc_target, input_dict1, list_dict]
+ )
pd.testing.assert_frame_equal(expected_contrib, original)
@@ -90,61 +113,83 @@ def test_multiple_encoding_columntransfomers(self):
"""
Test multiple preprocessing columntransformers
"""
- train = pd.DataFrame({'Onehot1': ['A', 'B'], 'Onehot2': ['C', 'D'],
- 'Binary1': ['E', 'F'], 'Binary2': ['G', 'H'],
- 'Ordinal1': ['I', 'J'], 'Ordinal2': ['K', 'L'],
- 'BaseN1': ['M', 'N'], 'BaseN2': ['O', 'P'],
- 'Target1': ['Q', 'R'], 'Target2': ['S', 'T'],
- 'other': ['other', np.nan]})
-
- contributions = pd.DataFrame([[1, 0, 1, 1, 1, 0, 1, 1, 3, 0, -3.5, 0, 4, 4, 5, 5, 0, 6, 7, 0, 8, 9, 10],
- [.5, .5, 2, 0, .5, .5, 2, 0, 1.5, 1.5, 5.5, -2, -4, -4, -5, -5, 8.5, -2.5, -7, 14, -8, -9, -10]],
- index=['index1', 'index2'])
-
- expected_contrib = pd.DataFrame({'onehot_skp_Onehot1': [1., 1.], 'onehot_skp_Onehot2': [2, 2],
- 'onehot_ce_Onehot1': [1., 1.], 'onehot_ce_Onehot2': [2, 2],
- 'binary_ce_Binary1': [3., 3.], 'binary_ce_Binary2': [-3.5, 3.5],
- 'ordinal_ce_Ordinal1': [4, -4], 'ordinal_ce_Ordinal2': [4, -4],
- 'ordinal_skp_Ordinal1': [5, -5], 'ordinal_skp_Ordinal2': [5, -5],
- 'basen_ce_BaseN1': [6., 6.], 'basen_ce_BaseN2': [7, 7],
- 'target_ce_Target1': [8, -8], 'target_ce_Target2': [9, -9],
- 22: [10, -10]},
- index=['index1', 'index2'])
-
- y = pd.DataFrame(data=[0, 1], columns=['y'])
+ train = pd.DataFrame(
+ {
+ "Onehot1": ["A", "B"],
+ "Onehot2": ["C", "D"],
+ "Binary1": ["E", "F"],
+ "Binary2": ["G", "H"],
+ "Ordinal1": ["I", "J"],
+ "Ordinal2": ["K", "L"],
+ "BaseN1": ["M", "N"],
+ "BaseN2": ["O", "P"],
+ "Target1": ["Q", "R"],
+ "Target2": ["S", "T"],
+ "other": ["other", np.nan],
+ }
+ )
+
+ contributions = pd.DataFrame(
+ [
+ [1, 0, 1, 1, 1, 0, 1, 1, 3, 0, -3.5, 0, 4, 4, 5, 5, 0, 6, 7, 0, 8, 9, 10],
+ [0.5, 0.5, 2, 0, 0.5, 0.5, 2, 0, 1.5, 1.5, 5.5, -2, -4, -4, -5, -5, 8.5, -2.5, -7, 14, -8, -9, -10],
+ ],
+ index=["index1", "index2"],
+ )
+
+ expected_contrib = pd.DataFrame(
+ {
+ "onehot_skp_Onehot1": [1.0, 1.0],
+ "onehot_skp_Onehot2": [2, 2],
+ "onehot_ce_Onehot1": [1.0, 1.0],
+ "onehot_ce_Onehot2": [2, 2],
+ "binary_ce_Binary1": [3.0, 3.0],
+ "binary_ce_Binary2": [-3.5, 3.5],
+ "ordinal_ce_Ordinal1": [4, -4],
+ "ordinal_ce_Ordinal2": [4, -4],
+ "ordinal_skp_Ordinal1": [5, -5],
+ "ordinal_skp_Ordinal2": [5, -5],
+ "basen_ce_BaseN1": [6.0, 6.0],
+ "basen_ce_BaseN2": [7, 7],
+ "target_ce_Target1": [8, -8],
+ "target_ce_Target2": [9, -9],
+ 22: [10, -10],
+ },
+ index=["index1", "index2"],
+ )
+
+ y = pd.DataFrame(data=[0, 1], columns=["y"])
enc = ColumnTransformer(
transformers=[
- ('onehot_skp', skp.OneHotEncoder(), ['Onehot1', 'Onehot2']),
- ('onehot_ce', ce.OneHotEncoder(), ['Onehot1', 'Onehot2']),
- ('binary_ce', ce.BinaryEncoder(), ['Binary1', 'Binary2']),
- ('ordinal_ce', ce.OrdinalEncoder(), ['Ordinal1', 'Ordinal2']),
- ('ordinal_skp', skp.OrdinalEncoder(), ['Ordinal1', 'Ordinal2']),
- ('basen_ce', ce.BaseNEncoder(), ['BaseN1', 'BaseN2']),
- ('target_ce', ce.TargetEncoder(), ['Target1', 'Target2'])
+ ("onehot_skp", skp.OneHotEncoder(), ["Onehot1", "Onehot2"]),
+ ("onehot_ce", ce.OneHotEncoder(), ["Onehot1", "Onehot2"]),
+ ("binary_ce", ce.BinaryEncoder(), ["Binary1", "Binary2"]),
+ ("ordinal_ce", ce.OrdinalEncoder(), ["Ordinal1", "Ordinal2"]),
+ ("ordinal_skp", skp.OrdinalEncoder(), ["Ordinal1", "Ordinal2"]),
+ ("basen_ce", ce.BaseNEncoder(), ["BaseN1", "BaseN2"]),
+ ("target_ce", ce.TargetEncoder(), ["Target1", "Target2"]),
],
- remainder='passthrough')
+ remainder="passthrough",
+ )
enc.fit(train, y)
input_dict1 = dict()
- input_dict1['col'] = 'Onehot2'
- input_dict1['mapping'] = pd.Series(data=['C', 'D', np.nan], index=['C', 'D', 'missing'])
- input_dict1['data_type'] = 'object'
+ input_dict1["col"] = "Onehot2"
+ input_dict1["mapping"] = pd.Series(data=["C", "D", np.nan], index=["C", "D", "missing"])
+ input_dict1["data_type"] = "object"
input_dict2 = dict()
- input_dict2['col'] = 'Binary2'
- input_dict2['mapping'] = pd.Series(data=['G', 'H', np.nan], index=['G', 'H', 'missing'])
- input_dict2['data_type'] = 'object'
+ input_dict2["col"] = "Binary2"
+ input_dict2["mapping"] = pd.Series(data=["G", "H", np.nan], index=["G", "H", "missing"])
+ input_dict2["data_type"] = "object"
input_dict3 = dict()
- input_dict3['col'] = 'Ordinal2'
- input_dict3['mapping'] = pd.Series(data=['K', 'L', np.nan], index=['K', 'L', 'missing'])
- input_dict3['data_type'] = 'object'
+ input_dict3["col"] = "Ordinal2"
+ input_dict3["mapping"] = pd.Series(data=["K", "L", np.nan], index=["K", "L", "missing"])
+ input_dict3["data_type"] = "object"
list_dict = [input_dict2, input_dict3]
- original = inverse_transform_contributions(contributions,
- [enc,
- input_dict1,
- list_dict])
+ original = inverse_transform_contributions(contributions, [enc, input_dict1, list_dict])
- pd.testing.assert_frame_equal(expected_contrib, original)
\ No newline at end of file
+ pd.testing.assert_frame_equal(expected_contrib, original)
diff --git a/tests/unit_tests/explainer/test_consistency.py b/tests/unit_tests/explainer/test_consistency.py
index ec6d215d..13cc6b9a 100644
--- a/tests/unit_tests/explainer/test_consistency.py
+++ b/tests/unit_tests/explainer/test_consistency.py
@@ -1,8 +1,10 @@
import itertools
+import unittest
+
import numpy as np
import pandas as pd
from sklearn.ensemble import RandomForestClassifier
-import unittest
+
from shapash.explainer.consistency import Consistency
@@ -10,37 +12,31 @@ class TestConsistency(unittest.TestCase):
"""
Unit test consistency
"""
+
def setUp(self):
self.df = pd.DataFrame(
- data=np.array([[1, 2, 3, 0],
- [2, 4, 6, 1],
- [2, 2, 1, 0],
- [2, 4, 1, 1]]),
- columns=['X1', 'X2', 'X3', 'y'])
+ data=np.array([[1, 2, 3, 0], [2, 4, 6, 1], [2, 2, 1, 0], [2, 4, 1, 1]]), columns=["X1", "X2", "X3", "y"]
+ )
self.X = self.df.iloc[:, :-1]
self.y = self.df.iloc[:, -1]
- self.w1 = pd.DataFrame(np.array([[0.14, 0.04, 0.17],
- [0.02, 0.01, 0.33],
- [0.56, 0.12, 0.29],
- [0.03, 0.01, 0.04]]),
- columns=['X1', 'X2', 'X3'])
- self.w2 = pd.DataFrame(np.array([[0.38, 0.35, 0.01],
- [0.01, 0.30, 0.05],
- [0.45, 0.41, 0.12],
- [0.07, 0.30, 0.21]]),
- columns=['X1', 'X2', 'X3'])
- self.w3 = pd.DataFrame(np.array([[0.49, 0.17, 0.02],
- [0.25, 0.12, 0.25],
- [0.01, 0.06, 0.06],
- [0.19, 0.02, 0.18]]),
- columns=['X1', 'X2', 'X3'])
+ self.w1 = pd.DataFrame(
+ np.array([[0.14, 0.04, 0.17], [0.02, 0.01, 0.33], [0.56, 0.12, 0.29], [0.03, 0.01, 0.04]]),
+ columns=["X1", "X2", "X3"],
+ )
+ self.w2 = pd.DataFrame(
+ np.array([[0.38, 0.35, 0.01], [0.01, 0.30, 0.05], [0.45, 0.41, 0.12], [0.07, 0.30, 0.21]]),
+ columns=["X1", "X2", "X3"],
+ )
+ self.w3 = pd.DataFrame(
+ np.array([[0.49, 0.17, 0.02], [0.25, 0.12, 0.25], [0.01, 0.06, 0.06], [0.19, 0.02, 0.18]]),
+ columns=["X1", "X2", "X3"],
+ )
self.contributions = {"contrib_1": self.w1, "contrib_2": self.w2, "contrib_3": self.w3}
self.cns = Consistency()
- self.cns.compile(contributions = self.contributions, x=self.X)
-
+ self.cns.compile(contributions=self.contributions, x=self.X)
def test_compile(self):
assert isinstance(self.cns.methods, list)
@@ -49,34 +45,33 @@ def test_compile(self):
assert self.cns.weights[0].shape == self.w1.shape
assert all(x.shape == self.cns.weights[0].shape for x in self.cns.weights)
-
def test_check_consistency_contributions(self):
weights = [self.w1, self.w2, self.w3]
if weights[0].ndim == 1:
- raise ValueError('Multiple datapoints are required to compute the metric')
+ raise ValueError("Multiple datapoints are required to compute the metric")
if not all(isinstance(x, pd.DataFrame) for x in weights):
- raise ValueError('Contributions must be pandas DataFrames')
+ raise ValueError("Contributions must be pandas DataFrames")
if not all(x.shape == weights[0].shape for x in weights):
- raise ValueError('Contributions must be of same shape')
+ raise ValueError("Contributions must be of same shape")
if not all(x.columns.tolist() == weights[0].columns.tolist() for x in weights):
- raise ValueError('Columns names are different between contributions')
+ raise ValueError("Columns names are different between contributions")
if not all(x.index.tolist() == weights[0].index.tolist() for x in weights):
- raise ValueError('Index names are different between contributions')
+ raise ValueError("Index names are different between contributions")
def test_calculate_all_distances(self):
all_comparisons, mean_distances = self.cns.calculate_all_distances(self.cns.methods, self.cns.weights)
num_comb = len(list(itertools.combinations(self.cns.methods, 2)))
- assert all_comparisons.shape == (num_comb*self.X.shape[0], 4)
+ assert all_comparisons.shape == (num_comb * self.X.shape[0], 4)
assert isinstance(mean_distances, pd.DataFrame)
assert mean_distances.shape == (len(self.cns.methods), len(self.cns.methods))
def test_calculate_pairwise_distances(self):
l2_dist = self.cns.calculate_pairwise_distances(self.cns.weights, 0, 1)
- assert l2_dist.shape == (self.X.shape[0], )
+ assert l2_dist.shape == (self.X.shape[0],)
def test_calculate_mean_distances(self):
_, mean_distances = self.cns.calculate_all_distances(self.cns.methods, self.cns.weights)
@@ -85,8 +80,10 @@ def test_calculate_mean_distances(self):
self.cns.calculate_mean_distances(self.cns.methods, mean_distances, 0, 1, l2_dist)
assert mean_distances.shape == (len(self.cns.methods), len(self.cns.methods))
- assert mean_distances.loc[self.cns.methods[0], self.cns.methods[1]] == \
- mean_distances.loc[self.cns.methods[1], self.cns.methods[0]]
+ assert (
+ mean_distances.loc[self.cns.methods[0], self.cns.methods[1]]
+ == mean_distances.loc[self.cns.methods[1], self.cns.methods[0]]
+ )
def test_find_examples(self):
weights = [weight.values for weight in self.cns.weights]
@@ -109,9 +106,7 @@ def test_pairwise_consistency_plot(self):
methods = ["contrib_1", "contrib_3"]
max_features = 2
max_points = 100
- output = self.cns.pairwise_consistency_plot(methods=methods,
- max_features=max_features,
- max_points=max_points)
+ output = self.cns.pairwise_consistency_plot(methods=methods, max_features=max_features, max_points=max_points)
assert len(output.data[0].x) == min(max_points, len(self.X))
assert len(output.data) == 2 * max_features + 1
diff --git a/tests/unit_tests/explainer/test_multi_decorator.py b/tests/unit_tests/explainer/test_multi_decorator.py
index a9a37008..f2b03a59 100644
--- a/tests/unit_tests/explainer/test_multi_decorator.py
+++ b/tests/unit_tests/explainer/test_multi_decorator.py
@@ -2,9 +2,11 @@
Unit test for multi decorator
"""
import unittest
-from unittest.mock import patch, Mock
+from unittest.mock import Mock, patch
+
import numpy as np
import pandas as pd
+
from shapash.explainer.multi_decorator import MultiDecorator
@@ -32,7 +34,7 @@ def dummy_function(self, xparam, yparam):
[type]
[description]
"""
- return '+'.join([str(xparam), str(yparam)])
+ return "+".join([str(xparam), str(yparam)])
class TestMultiDecorator(unittest.TestCase):
@@ -40,13 +42,14 @@ class TestMultiDecorator(unittest.TestCase):
Unit test multi decorator Class
TODO: Docstring
"""
+
def test_delegate_1(self):
"""
Unit test delegate 1
"""
state = MultiDecorator(DummyState())
- output = state.delegate('dummy_function', [1, 2, 3], 9)
- expected = ['1+9', '2+9', '3+9']
+ output = state.delegate("dummy_function", [1, 2, 3], 9)
+ expected = ["1+9", "2+9", "3+9"]
self.assertListEqual(output, expected)
def test_delegate_2(self):
@@ -55,7 +58,7 @@ def test_delegate_2(self):
"""
state = MultiDecorator(DummyState())
with self.assertRaises(ValueError):
- state.delegate('dummy_function')
+ state.delegate("dummy_function")
def test_delegate_3(self):
"""
@@ -63,7 +66,7 @@ def test_delegate_3(self):
"""
state = MultiDecorator(DummyState())
with self.assertRaises(AttributeError):
- state.delegate('non_existing_function', [1, 2, 3], 9)
+ state.delegate("non_existing_function", [1, 2, 3], 9)
def test_delegate_4(self):
"""
@@ -71,7 +74,7 @@ def test_delegate_4(self):
"""
state = MultiDecorator(DummyState())
with self.assertRaises(ValueError):
- state.delegate('dummy_member', [1, 2, 3], 9)
+ state.delegate("dummy_member", [1, 2, 3], 9)
def test_delegate_5(self):
"""
@@ -79,7 +82,7 @@ def test_delegate_5(self):
"""
state = MultiDecorator(DummyState())
with self.assertRaises(ValueError):
- state.delegate('dummy_function', 1, 9)
+ state.delegate("dummy_function", 1, 9)
def test_assign_contributions(self):
"""
@@ -88,17 +91,14 @@ def test_assign_contributions(self):
backend = Mock()
backend.assign_contributions = Mock(
side_effect=[
- {'a': 1, 'b': 2},
- {'a': 3, 'b': 4},
- {'a': 5, 'b': 6},
+ {"a": 1, "b": 2},
+ {"a": 3, "b": 4},
+ {"a": 5, "b": 6},
]
)
state = MultiDecorator(backend)
output = state.assign_contributions([1, 2, 3])
- expected = {
- 'a': [1, 3, 5],
- 'b': [2, 4, 6]
- }
+ expected = {"a": [1, 3, 5], "b": [2, 4, 6]}
self.assertDictEqual(output, expected)
def test_check_contributions_1(self):
@@ -121,7 +121,7 @@ def test_check_contributions_2(self):
output = state.check_contributions([1, 2, 3], Mock())
assert not output
- @patch('shapash.explainer.multi_decorator.MultiDecorator.delegate')
+ @patch("shapash.explainer.multi_decorator.MultiDecorator.delegate")
def test_combine_masks(self, mock_delegate):
"""
Unit test combine masks
@@ -143,7 +143,7 @@ def test_combine_masks(self, mock_delegate):
backend.combine_masks = Mock()
state = MultiDecorator(backend)
state.combine_masks(masks)
- mock_delegate.assert_called_with('combine_masks', transposed_masks)
+ mock_delegate.assert_called_with("combine_masks", transposed_masks)
def test_compute_masked_contributions(self):
"""
@@ -174,12 +174,12 @@ def test_summarize_1(self):
var_dict1 = pd.DataFrame([[1, 0, 2, 3], [1, 0, 3, 2]])
var_dict2 = pd.DataFrame([[1, 0, 2, 3], [1, 0, 3, 2]])
var_dict = [var_dict1, var_dict2]
- x_sorted1 = pd.DataFrame([[1., 3., 22., 1.], [2., 1., 2., 38.]])
- x_sorted2 = pd.DataFrame([[1., 3., 22., 1.], [2., 1., 2., 38.]])
+ x_sorted1 = pd.DataFrame([[1.0, 3.0, 22.0, 1.0], [2.0, 1.0, 2.0, 38.0]])
+ x_sorted2 = pd.DataFrame([[1.0, 3.0, 22.0, 1.0], [2.0, 1.0, 2.0, 38.0]])
x_sorted = [x_sorted1, x_sorted2]
mask = pd.DataFrame([[True, True, False, False], [True, True, False, False]])
- columns_dict = {0: 'Pclass', 1: 'Sex', 2: 'Age', 3: 'Embarked'}
- features_dict = {'Pclass': 'Pclass', 'Sex': 'Sex', 'Age': 'Age (years)', 'Embarked': 'emb'}
+ columns_dict = {0: "Pclass", 1: "Sex", 2: "Age", 3: "Embarked"}
+ features_dict = {"Pclass": "Pclass", "Sex": "Sex", "Age": "Age (years)", "Embarked": "emb"}
state = MultiDecorator(backend)
state.summarize(contrib_sorted, var_dict, x_sorted, mask, columns_dict, features_dict)
assert backend.summarize.call_count == 2
diff --git a/tests/unit_tests/explainer/test_smart_explainer.py b/tests/unit_tests/explainer/test_smart_explainer.py
index dff25970..97269c09 100644
--- a/tests/unit_tests/explainer/test_smart_explainer.py
+++ b/tests/unit_tests/explainer/test_smart_explainer.py
@@ -1,27 +1,30 @@
"""
Unit test for smart explainer
"""
-import unittest
-from unittest.mock import patch, Mock
import os
import sys
+import types
+import unittest
from os import path
from pathlib import Path
-import types
-import pandas as pd
-import numpy as np
+from unittest.mock import Mock, patch
+
import catboost as cb
+import category_encoders as ce
+import numpy as np
+import pandas as pd
+import shap
+from catboost import CatBoostClassifier, CatBoostRegressor
from pandas.testing import assert_frame_equal
-from sklearn.tree import DecisionTreeRegressor, DecisionTreeClassifier
from sklearn.ensemble import RandomForestClassifier
-from catboost import CatBoostClassifier, CatBoostRegressor
+from sklearn.tree import DecisionTreeClassifier, DecisionTreeRegressor
+
from shapash import SmartExplainer
-from shapash.explainer.multi_decorator import MultiDecorator
from shapash.backend import ShapBackend
+from shapash.explainer.multi_decorator import MultiDecorator
from shapash.explainer.smart_state import SmartState
-import category_encoders as ce
from shapash.utils.check import check_model
-import shap
+
def init_sme_to_pickle_test():
"""
@@ -33,9 +36,9 @@ def init_sme_to_pickle_test():
[description]
"""
current = Path(path.abspath(__file__)).parent.parent.parent
- pkl_file = path.join(current, 'data/xpl.pkl')
+ pkl_file = path.join(current, "data/xpl.pkl")
contributions = pd.DataFrame([[-0.1, 0.2, -0.3], [0.1, -0.2, 0.3]])
- y_pred = pd.DataFrame(data=np.array([1, 2]), columns=['pred'])
+ y_pred = pd.DataFrame(data=np.array([1, 2]), columns=["pred"])
dataframe_x = pd.DataFrame([[1, 2, 3], [1, 2, 3]])
model = DecisionTreeRegressor().fit(dataframe_x, y_pred)
xpl = SmartExplainer(model=model)
@@ -43,6 +46,7 @@ def init_sme_to_pickle_test():
xpl.filter(max_contrib=2)
return pkl_file, xpl
+
class TestSmartExplainer(unittest.TestCase):
"""
Unit test smart explainer
@@ -50,12 +54,7 @@ class TestSmartExplainer(unittest.TestCase):
"""
def setUp(self) -> None:
- x_init = pd.DataFrame(
- [[1, 2],
- [3, 4]],
- columns=['Col1', 'Col2'],
- index=['Id1', 'Id2']
- )
+ x_init = pd.DataFrame([[1, 2], [3, 4]], columns=["Col1", "Col2"], index=["Id1", "Id2"])
self.model = CatBoostRegressor().fit(x_init, [0, 1])
def test_init(self):
@@ -63,36 +62,27 @@ def test_init(self):
test init smart explainer
"""
xpl = SmartExplainer(self.model)
- assert hasattr(xpl, 'plot')
+ assert hasattr(xpl, "plot")
def assertRaisesWithMessage(self, msg, func, *args, **kwargs):
try:
func(*args, **kwargs)
self.assertFail()
except Exception as inst:
- self.assertEqual(inst.args[0]['message'], msg)
+ self.assertEqual(inst.args[0]["message"], msg)
def test_modify_postprocessing_1(self):
"""
Unit test modify postprocessing 1
"""
xpl = SmartExplainer(self.model)
- xpl.x_init = pd.DataFrame(
- [[1, 2],
- [3, 4]],
- columns=['Col1', 'Col2'],
- index=['Id1', 'Id2']
- )
- xpl.features_dict = {'Col1': 'Column1', 'Col2': 'Column2'}
- xpl.columns_dict = {0: 'Col1', 1:'Col2'}
- xpl.inv_features_dict = {'Column1': 'Col1', 'Column2': 'Col2'}
- postprocessing = {0: {'type' : 'suffix', 'rule':' t'},
- 'Column2': {'type' : 'prefix', 'rule' : 'test'}}
-
- expected_output = {
- 'Col1': {'type' : 'suffix', 'rule':' t'},
- 'Col2': {'type' : 'prefix', 'rule' : 'test'}
- }
+ xpl.x_init = pd.DataFrame([[1, 2], [3, 4]], columns=["Col1", "Col2"], index=["Id1", "Id2"])
+ xpl.features_dict = {"Col1": "Column1", "Col2": "Column2"}
+ xpl.columns_dict = {0: "Col1", 1: "Col2"}
+ xpl.inv_features_dict = {"Column1": "Col1", "Column2": "Col2"}
+ postprocessing = {0: {"type": "suffix", "rule": " t"}, "Column2": {"type": "prefix", "rule": "test"}}
+
+ expected_output = {"Col1": {"type": "suffix", "rule": " t"}, "Col2": {"type": "prefix", "rule": "test"}}
output = xpl.modify_postprocessing(postprocessing)
assert output == expected_output
@@ -101,16 +91,11 @@ def test_modify_postprocessing_2(self):
Unit test modify postprocessing 2
"""
xpl = SmartExplainer(self.model)
- xpl.x_init = pd.DataFrame(
- [[1, 2],
- [3, 4]],
- columns=['Col1', 'Col2'],
- index=['Id1', 'Id2']
- )
- xpl.features_dict = {'Col1': 'Column1', 'Col2': 'Column2'}
- xpl.columns_dict = {0: 'Col1', 1: 'Col2'}
- xpl.inv_features_dict = {'Column1': 'Col1', 'Column2': 'Col2'}
- postprocessing = {'Error': {'type': 'suffix', 'rule': ' t'}}
+ xpl.x_init = pd.DataFrame([[1, 2], [3, 4]], columns=["Col1", "Col2"], index=["Id1", "Id2"])
+ xpl.features_dict = {"Col1": "Column1", "Col2": "Column2"}
+ xpl.columns_dict = {0: "Col1", 1: "Col2"}
+ xpl.inv_features_dict = {"Column1": "Col1", "Column2": "Col2"}
+ postprocessing = {"Error": {"type": "suffix", "rule": " t"}}
with self.assertRaises(ValueError):
xpl.modify_postprocessing(postprocessing)
@@ -119,15 +104,10 @@ def test_apply_postprocessing_1(self):
Unit test apply_postprocessing 1
"""
xpl = SmartExplainer(self.model)
- xpl.x_init = pd.DataFrame(
- [[1, 2],
- [3, 4]],
- columns=['Col1', 'Col2'],
- index=['Id1', 'Id2']
- )
- xpl.features_dict = {'Col1': 'Column1', 'Col2': 'Column2'}
- xpl.columns_dict = {0: 'Col1', 1: 'Col2'}
- xpl.inv_features_dict = {'Column1': 'Col1', 'Column2': 'Col2'}
+ xpl.x_init = pd.DataFrame([[1, 2], [3, 4]], columns=["Col1", "Col2"], index=["Id1", "Id2"])
+ xpl.features_dict = {"Col1": "Column1", "Col2": "Column2"}
+ xpl.columns_dict = {0: "Col1", 1: "Col2"}
+ xpl.inv_features_dict = {"Column1": "Col1", "Column2": "Col2"}
assert np.array_equal(xpl.x_init, xpl.apply_postprocessing())
def test_apply_postprocessing_2(self):
@@ -135,22 +115,13 @@ def test_apply_postprocessing_2(self):
Unit test apply_postprocessing 2
"""
xpl = SmartExplainer(self.model)
- xpl.x_init = pd.DataFrame(
- [[1, 2],
- [3, 4]],
- columns=['Col1', 'Col2'],
- index=['Id1', 'Id2']
- )
- xpl.features_dict = {'Col1': 'Column1', 'Col2': 'Column2'}
- xpl.columns_dict = {0: 'Col1', 1: 'Col2'}
- xpl.inv_features_dict = {'Column1': 'Col1', 'Column2': 'Col2'}
- postprocessing = {'Col1': {'type': 'suffix', 'rule': ' t'},
- 'Col2': {'type': 'prefix', 'rule': 'test'}}
+ xpl.x_init = pd.DataFrame([[1, 2], [3, 4]], columns=["Col1", "Col2"], index=["Id1", "Id2"])
+ xpl.features_dict = {"Col1": "Column1", "Col2": "Column2"}
+ xpl.columns_dict = {0: "Col1", 1: "Col2"}
+ xpl.inv_features_dict = {"Column1": "Col1", "Column2": "Col2"}
+ postprocessing = {"Col1": {"type": "suffix", "rule": " t"}, "Col2": {"type": "prefix", "rule": "test"}}
expected_output = pd.DataFrame(
- data=[['1 t', 'test2'],
- ['3 t', 'test4']],
- columns=['Col1', 'Col2'],
- index=['Id1', 'Id2']
+ data=[["1 t", "test2"], ["3 t", "test4"]], columns=["Col1", "Col2"], index=["Id1", "Id2"]
)
output = xpl.apply_postprocessing(postprocessing)
assert np.array_equal(output, expected_output)
@@ -181,9 +152,9 @@ def test_check_label_dict_1(self):
"""
Unit test check label dict 1
"""
- xpl = SmartExplainer(self.model, label_dict={1: 'Yes', 0: 'No'})
+ xpl = SmartExplainer(self.model, label_dict={1: "Yes", 0: "No"})
xpl._classes = [0, 1]
- xpl._case = 'classification'
+ xpl._case = "classification"
xpl.check_label_dict()
def test_check_label_dict_2(self):
@@ -191,32 +162,32 @@ def test_check_label_dict_2(self):
Unit test check label dict 2
"""
xpl = SmartExplainer(self.model)
- xpl._case = 'regression'
+ xpl._case = "regression"
xpl.check_label_dict()
def test_check_features_dict_1(self):
"""
Unit test check features dict 1
"""
- xpl = SmartExplainer(self.model, features_dict={'Age': 'Age (Years Old)'})
- xpl.columns_dict = {0: 'Age', 1: 'Education', 2: 'Sex'}
+ xpl = SmartExplainer(self.model, features_dict={"Age": "Age (Years Old)"})
+ xpl.columns_dict = {0: "Age", 1: "Education", 2: "Sex"}
xpl.check_features_dict()
- assert xpl.features_dict['Age'] == 'Age (Years Old)'
- assert xpl.features_dict['Education'] == 'Education'
+ assert xpl.features_dict["Age"] == "Age (Years Old)"
+ assert xpl.features_dict["Education"] == "Education"
def test_compile_1(self):
"""
Unit test compile 1
checking compile method without model
"""
- df = pd.DataFrame(range(0, 21), columns=['id'])
- df['y'] = df['id'].apply(lambda x: 1 if x < 10 else 0)
- df['x1'] = np.random.randint(1, 123, df.shape[0])
- df['x2'] = np.random.randint(1, 3, df.shape[0])
- df = df.set_index('id')
- clf = cb.CatBoostClassifier(n_estimators=1).fit(df[['x1', 'x2']], df['y'])
+ df = pd.DataFrame(range(0, 21), columns=["id"])
+ df["y"] = df["id"].apply(lambda x: 1 if x < 10 else 0)
+ df["x1"] = np.random.randint(1, 123, df.shape[0])
+ df["x2"] = np.random.randint(1, 3, df.shape[0])
+ df = df.set_index("id")
+ clf = cb.CatBoostClassifier(n_estimators=1).fit(df[["x1", "x2"]], df["y"])
xpl = SmartExplainer(clf)
- xpl.compile(x=df[['x1', 'x2']])
+ xpl.compile(x=df[["x1", "x2"]])
assert xpl._case == "classification"
self.assertListEqual(xpl._classes, [0, 1])
@@ -225,39 +196,35 @@ def test_compile_2(self):
Unit test compile 2
checking new attributes added to the compile method
"""
- df = pd.DataFrame(range(0, 5), columns=['id'])
- df['y'] = df['id'].apply(lambda x: 1 if x < 2 else 0)
- df['x1'] = np.random.randint(1, 123, df.shape[0])
- df['x2'] = ["S", "M", "S", "D", "M"]
- df = df.set_index('id')
+ df = pd.DataFrame(range(0, 5), columns=["id"])
+ df["y"] = df["id"].apply(lambda x: 1 if x < 2 else 0)
+ df["x1"] = np.random.randint(1, 123, df.shape[0])
+ df["x2"] = ["S", "M", "S", "D", "M"]
+ df = df.set_index("id")
encoder = ce.OrdinalEncoder(cols=["x2"], handle_unknown="None")
encoder_fitted = encoder.fit(df)
df_encoded = encoder_fitted.transform(df)
output = df[["x1", "x2"]].copy()
output["x2"] = ["single", "married", "single", "divorced", "married"]
- clf = cb.CatBoostClassifier(n_estimators=1).fit(df_encoded[['x1', 'x2']], df_encoded['y'])
+ clf = cb.CatBoostClassifier(n_estimators=1).fit(df_encoded[["x1", "x2"]], df_encoded["y"])
- postprocessing_1 = {"x2": {
- "type": "transcoding",
- "rule": {"S": "single", "M": "married", "D": "divorced"}}}
+ postprocessing_1 = {"x2": {"type": "transcoding", "rule": {"S": "single", "M": "married", "D": "divorced"}}}
postprocessing_2 = {
- "family_situation": {
- "type": "transcoding",
- "rule": {"S": "single", "M": "married", "D": "divorced"}}}
-
- xpl_postprocessing1 = SmartExplainer(clf, preprocessing=encoder_fitted,
- postprocessing=postprocessing_1)
- xpl_postprocessing2 = SmartExplainer(clf, features_dict={"x1": "age",
- "x2": "family_situation"},
- preprocessing=encoder_fitted,
- postprocessing=postprocessing_2
- )
- xpl_postprocessing3 = SmartExplainer(clf, preprocessing=None,
- postprocessing=None)
-
- xpl_postprocessing1.compile(x=df_encoded[['x1', 'x2']])
- xpl_postprocessing2.compile(x=df_encoded[['x1', 'x2']])
- xpl_postprocessing3.compile(x=df_encoded[['x1', 'x2']])
+ "family_situation": {"type": "transcoding", "rule": {"S": "single", "M": "married", "D": "divorced"}}
+ }
+
+ xpl_postprocessing1 = SmartExplainer(clf, preprocessing=encoder_fitted, postprocessing=postprocessing_1)
+ xpl_postprocessing2 = SmartExplainer(
+ clf,
+ features_dict={"x1": "age", "x2": "family_situation"},
+ preprocessing=encoder_fitted,
+ postprocessing=postprocessing_2,
+ )
+ xpl_postprocessing3 = SmartExplainer(clf, preprocessing=None, postprocessing=None)
+
+ xpl_postprocessing1.compile(x=df_encoded[["x1", "x2"]])
+ xpl_postprocessing2.compile(x=df_encoded[["x1", "x2"]])
+ xpl_postprocessing3.compile(x=df_encoded[["x1", "x2"]])
assert hasattr(xpl_postprocessing1, "preprocessing")
assert hasattr(xpl_postprocessing1, "postprocessing")
@@ -277,24 +244,22 @@ def test_compile_3(self):
Unit test compile 3
checking compile method without model
"""
- df = pd.DataFrame(range(0, 21), columns=['id'])
- df['y'] = df['id'].apply(lambda x: 1 if x < 10 else 0)
- df['x1'] = np.random.randint(1, 123, df.shape[0])
- df['x2'] = np.random.randint(1, 3, df.shape[0])
- df = df.set_index('id')
- clf = cb.CatBoostClassifier(n_estimators=1).fit(df[['x1', 'x2']], df['y'])
+ df = pd.DataFrame(range(0, 21), columns=["id"])
+ df["y"] = df["id"].apply(lambda x: 1 if x < 10 else 0)
+ df["x1"] = np.random.randint(1, 123, df.shape[0])
+ df["x2"] = np.random.randint(1, 3, df.shape[0])
+ df = df.set_index("id")
+ clf = cb.CatBoostClassifier(n_estimators=1).fit(df[["x1", "x2"]], df["y"])
contrib = pd.DataFrame(
- [[1, 2, 3, 4],
- [5, 6, 7, 8],
- [9, 10, 11, 12]],
- columns=['contribution_0', 'contribution_1', 'contribution_2', 'contribution_3'],
- index=[0, 1, 2]
+ [[1, 2, 3, 4], [5, 6, 7, 8], [9, 10, 11, 12]],
+ columns=["contribution_0", "contribution_1", "contribution_2", "contribution_3"],
+ index=[0, 1, 2],
)
xpl = SmartExplainer(clf)
with self.assertRaises(ValueError):
- xpl.compile(x=df[['x1', 'x2']], contributions=contrib)
+ xpl.compile(x=df[["x1", "x2"]], contributions=contrib)
def test_compile_4(self):
"""
@@ -302,31 +267,31 @@ def test_compile_4(self):
checking compile method with lime backend
"""
np.random.seed(1)
- df = pd.DataFrame(range(0, 5), columns=['id'])
- df['y'] = df['id'].apply(lambda x: 1 if x < 2 else 0)
- df['x1'] = np.random.randint(1, 123, df.shape[0])
- df['x2'] = np.random.randint(1, 3, df.shape[0])
- df = df.set_index('id')
- clf = RandomForestClassifier(n_estimators=1).fit(df[['x1', 'x2']], df['y'])
+ df = pd.DataFrame(range(0, 5), columns=["id"])
+ df["y"] = df["id"].apply(lambda x: 1 if x < 2 else 0)
+ df["x1"] = np.random.randint(1, 123, df.shape[0])
+ df["x2"] = np.random.randint(1, 3, df.shape[0])
+ df = df.set_index("id")
+ clf = RandomForestClassifier(n_estimators=1).fit(df[["x1", "x2"]], df["y"])
- xpl = SmartExplainer(clf, data=df[['x1', 'x2']], backend="lime")
- xpl.compile(x=df[['x1', 'x2']])
+ xpl = SmartExplainer(clf, data=df[["x1", "x2"]], backend="lime")
+ xpl.compile(x=df[["x1", "x2"]])
def test_compile_5(self):
"""
Unit test compile 5
checking compile method with y_target
"""
- df = pd.DataFrame(range(0, 21), columns=['id'])
- df['y'] = df['id'].apply(lambda x: 1 if x < 10 else 0)
- df['x1'] = np.random.randint(1, 123, df.shape[0])
- df['x2'] = np.random.randint(1, 3, df.shape[0])
- df = df.set_index('id')
- clf = cb.CatBoostClassifier(n_estimators=1).fit(df[['x1', 'x2']], df['y'])
+ df = pd.DataFrame(range(0, 21), columns=["id"])
+ df["y"] = df["id"].apply(lambda x: 1 if x < 10 else 0)
+ df["x1"] = np.random.randint(1, 123, df.shape[0])
+ df["x2"] = np.random.randint(1, 3, df.shape[0])
+ df = df.set_index("id")
+ clf = cb.CatBoostClassifier(n_estimators=1).fit(df[["x1", "x2"]], df["y"])
xpl = SmartExplainer(clf)
- xpl.compile(x=df[['x1', 'x2']], y_target=df['y'])
+ xpl.compile(x=df[["x1", "x2"]], y_target=df["y"])
assert xpl._case == "classification"
- assert_frame_equal(xpl.y_target, df[['y']])
+ assert_frame_equal(xpl.y_target, df[["y"]])
self.assertListEqual(xpl._classes, [0, 1])
def test_compile_6(self):
@@ -335,24 +300,24 @@ def test_compile_6(self):
checking compile method with additional_data
"""
np.random.seed(1)
- df = pd.DataFrame(range(0, 5), columns=['id'])
- df['y'] = df['id'].apply(lambda x: 1 if x < 2 else 0)
- df['x1'] = np.random.randint(1, 123, df.shape[0])
- df['x2'] = np.random.randint(1, 3, df.shape[0])
- df['x3'] = np.random.randint(1, 3, df.shape[0])
- df = df.set_index('id')
- clf = RandomForestClassifier(n_estimators=1).fit(df[['x1', 'x2']], df['y'])
+ df = pd.DataFrame(range(0, 5), columns=["id"])
+ df["y"] = df["id"].apply(lambda x: 1 if x < 2 else 0)
+ df["x1"] = np.random.randint(1, 123, df.shape[0])
+ df["x2"] = np.random.randint(1, 3, df.shape[0])
+ df["x3"] = np.random.randint(1, 3, df.shape[0])
+ df = df.set_index("id")
+ clf = RandomForestClassifier(n_estimators=1).fit(df[["x1", "x2"]], df["y"])
xpl = SmartExplainer(clf)
- xpl.compile(x=df[['x1', 'x2']], additional_data=df[['x3']])
- assert len(xpl.additional_features_dict)==1
+ xpl.compile(x=df[["x1", "x2"]], additional_data=df[["x3"]])
+ assert len(xpl.additional_features_dict) == 1
def test_filter_0(self):
"""
Unit test filter 0
"""
xpl = SmartExplainer(self.model)
- mock_data = {'var_dict': 1, 'contrib_sorted': 2, 'x_sorted': 3}
+ mock_data = {"var_dict": 1, "contrib_sorted": 2, "x_sorted": 3}
xpl.data = mock_data
mockstate = Mock()
xpl.state = mockstate
@@ -363,11 +328,11 @@ def test_filter_0(self):
mockstate.sign_contributions.assert_not_called()
mockstate.combine_masks.assert_called()
mockstate.cutoff_contributions.assert_not_called()
- assert hasattr(xpl, 'mask')
+ assert hasattr(xpl, "mask")
mockstate.compute_masked_contributions.assert_called()
- assert hasattr(xpl, 'masked_contributions')
+ assert hasattr(xpl, "masked_contributions")
- @patch('shapash.explainer.smart_explainer.SmartExplainer.check_features_name')
+ @patch("shapash.explainer.smart_explainer.SmartExplainer.check_features_name")
def test_filter_1(self, mock_check_features_name):
"""
Unit test filter 1
@@ -378,27 +343,27 @@ def test_filter_1(self, mock_check_features_name):
"""
xpl = SmartExplainer(self.model)
mock_check_features_name.return_value = [1, 2]
- mock_data = {'var_dict': 1, 'contrib_sorted': 2, 'x_sorted': 3}
+ mock_data = {"var_dict": 1, "contrib_sorted": 2, "x_sorted": 3}
xpl.data = mock_data
mockstate = Mock()
xpl.state = mockstate
- xpl.filter(features_to_hide=['X1', 'X2'])
+ xpl.filter(features_to_hide=["X1", "X2"])
mockstate.init_mask.assert_called()
mockstate.hide_contributions.assert_called()
mockstate.cap_contributions.assert_not_called()
mockstate.sign_contributions.assert_not_called()
mockstate.combine_masks.assert_called()
mockstate.cutoff_contributions.assert_not_called()
- assert hasattr(xpl, 'mask')
+ assert hasattr(xpl, "mask")
mockstate.compute_masked_contributions.assert_called()
- assert hasattr(xpl, 'masked_contributions')
+ assert hasattr(xpl, "masked_contributions")
def test_filter_2(self):
"""
Unit test filter 2
"""
xpl = SmartExplainer(self.model)
- mock_data = {'var_dict': 1, 'contrib_sorted': 2, 'x_sorted': 3}
+ mock_data = {"var_dict": 1, "contrib_sorted": 2, "x_sorted": 3}
xpl.data = mock_data
mockstate = Mock()
xpl.state = mockstate
@@ -409,16 +374,16 @@ def test_filter_2(self):
mockstate.sign_contributions.assert_not_called()
mockstate.combine_masks.assert_called()
mockstate.cutoff_contributions.assert_not_called()
- assert hasattr(xpl, 'mask')
+ assert hasattr(xpl, "mask")
mockstate.compute_masked_contributions.assert_called()
- assert hasattr(xpl, 'masked_contributions')
+ assert hasattr(xpl, "masked_contributions")
def test_filter_3(self):
"""
Unit test filter 3
"""
xpl = SmartExplainer(self.model)
- mock_data = {'var_dict': 1, 'contrib_sorted': 2, 'x_sorted': 3}
+ mock_data = {"var_dict": 1, "contrib_sorted": 2, "x_sorted": 3}
xpl.data = mock_data
mockstate = Mock()
xpl.state = mockstate
@@ -429,16 +394,16 @@ def test_filter_3(self):
mockstate.sign_contributions.assert_called()
mockstate.combine_masks.assert_called()
mockstate.cutoff_contributions.assert_not_called()
- assert hasattr(xpl, 'mask')
+ assert hasattr(xpl, "mask")
mockstate.compute_masked_contributions.assert_called()
- assert hasattr(xpl, 'masked_contributions')
+ assert hasattr(xpl, "masked_contributions")
def test_filter_4(self):
"""
Unit test filter 4
"""
xpl = SmartExplainer(self.model)
- mock_data = {'var_dict': 1, 'contrib_sorted': 2, 'x_sorted': 3}
+ mock_data = {"var_dict": 1, "contrib_sorted": 2, "x_sorted": 3}
xpl.data = mock_data
mockstate = Mock()
xpl.state = mockstate
@@ -449,16 +414,16 @@ def test_filter_4(self):
mockstate.sign_contributions.assert_not_called()
mockstate.combine_masks.assert_called()
mockstate.cutoff_contributions.assert_called()
- assert hasattr(xpl, 'mask')
+ assert hasattr(xpl, "mask")
mockstate.compute_masked_contributions.assert_called()
- assert hasattr(xpl, 'masked_contributions')
+ assert hasattr(xpl, "masked_contributions")
def test_filter_5(self):
"""
Unit test filter 5
"""
xpl = SmartExplainer(self.model)
- mock_data = {'var_dict': 1, 'contrib_sorted': 2, 'x_sorted': 3}
+ mock_data = {"var_dict": 1, "contrib_sorted": 2, "x_sorted": 3}
xpl.data = mock_data
mockstate = Mock()
xpl.state = mockstate
@@ -469,16 +434,16 @@ def test_filter_5(self):
mockstate.sign_contributions.assert_called()
mockstate.combine_masks.assert_called()
mockstate.cutoff_contributions.assert_called()
- assert hasattr(xpl, 'mask')
+ assert hasattr(xpl, "mask")
mockstate.compute_masked_contributions.assert_called()
- assert hasattr(xpl, 'masked_contributions')
+ assert hasattr(xpl, "masked_contributions")
def test_filter_6(self):
"""
Unit test filter 6
"""
xpl = SmartExplainer(self.model)
- mock_data = {'var_dict': 1, 'contrib_sorted': 2, 'x_sorted': 3}
+ mock_data = {"var_dict": 1, "contrib_sorted": 2, "x_sorted": 3}
xpl.data = mock_data
mockstate = Mock()
xpl.state = mockstate
@@ -489,9 +454,9 @@ def test_filter_6(self):
mockstate.sign_contributions.assert_not_called()
mockstate.combine_masks.assert_called()
mockstate.cutoff_contributions.assert_not_called()
- assert hasattr(xpl, 'mask')
+ assert hasattr(xpl, "mask")
mockstate.compute_masked_contributions.assert_called()
- assert hasattr(xpl, 'masked_contributions')
+ assert hasattr(xpl, "masked_contributions")
def test_filter_7(self):
"""
@@ -499,66 +464,48 @@ def test_filter_7(self):
"""
xpl = SmartExplainer(self.model)
contributions = [
- pd.DataFrame(
- data=[[0.5, 0.4, 0.3], [0.9, 0.8, 0.7]],
- columns=['Col1', 'Col2', 'Col3']
- ),
- pd.DataFrame(
- data=[[0.3, 0.2, 0.1], [0.6, 0.5, 0.4]],
- columns=['Col1', 'Col2', 'Col3']
- )
+ pd.DataFrame(data=[[0.5, 0.4, 0.3], [0.9, 0.8, 0.7]], columns=["Col1", "Col2", "Col3"]),
+ pd.DataFrame(data=[[0.3, 0.2, 0.1], [0.6, 0.5, 0.4]], columns=["Col1", "Col2", "Col3"]),
]
- xpl.data = {'var_dict': 1, 'contrib_sorted': contributions, 'x_sorted': 3}
+ xpl.data = {"var_dict": 1, "contrib_sorted": contributions, "x_sorted": 3}
xpl.state = MultiDecorator(SmartState())
xpl.filter(threshold=0.5, max_contrib=2)
expected_mask = [
pd.DataFrame(
- data=[[True, False, False], [True, True, False]],
- columns=['contrib_1', 'contrib_2', 'contrib_3']
+ data=[[True, False, False], [True, True, False]], columns=["contrib_1", "contrib_2", "contrib_3"]
),
pd.DataFrame(
- data=[[False, False, False], [True, True, False]],
- columns=['contrib_1', 'contrib_2', 'contrib_3']
- )
+ data=[[False, False, False], [True, True, False]], columns=["contrib_1", "contrib_2", "contrib_3"]
+ ),
]
assert len(expected_mask) == len(xpl.mask)
test_list = [pd.testing.assert_frame_equal(e, m) for e, m in zip(expected_mask, xpl.mask)]
assert all(x is None for x in test_list)
expected_masked_contributions = [
- pd.DataFrame(
- data=[[0.0, 0.7], [0.0, 0.7]],
- columns=['masked_neg', 'masked_pos']
- ),
- pd.DataFrame(
- data=[[0.0, 0.6], [0.0, 0.4]],
- columns=['masked_neg', 'masked_pos']
- )
+ pd.DataFrame(data=[[0.0, 0.7], [0.0, 0.7]], columns=["masked_neg", "masked_pos"]),
+ pd.DataFrame(data=[[0.0, 0.6], [0.0, 0.4]], columns=["masked_neg", "masked_pos"]),
]
assert len(expected_masked_contributions) == len(xpl.masked_contributions)
- test_list = [pd.testing.assert_frame_equal(e, m) for e, m in
- zip(expected_masked_contributions, xpl.masked_contributions)]
+ test_list = [
+ pd.testing.assert_frame_equal(e, m) for e, m in zip(expected_masked_contributions, xpl.masked_contributions)
+ ]
assert all(x is None for x in test_list)
- expected_param_dict = {
- 'features_to_hide': None,
- 'threshold': 0.5,
- 'positive': None,
- 'max_contrib': 2
- }
+ expected_param_dict = {"features_to_hide": None, "threshold": 0.5, "positive": None, "max_contrib": 2}
self.assertDictEqual(expected_param_dict, xpl.mask_params)
def test_check_label_name_1(self):
"""
Unit test check label name 1
"""
- label_dict = {1: 'Age', 2: 'Education'}
+ label_dict = {1: "Age", 2: "Education"}
xpl = SmartExplainer(self.model, label_dict=label_dict)
xpl.inv_label_dict = {v: k for k, v in xpl.label_dict.items()}
xpl._classes = [1, 2]
- entry = 'Age'
+ entry = "Age"
expected_num = 0
expected_code = 1
- expected_value = 'Age'
- label_num, label_code, label_value = xpl.check_label_name(entry, 'value')
+ expected_value = "Age"
+ label_num, label_code, label_value = xpl.check_label_name(entry, "value")
assert expected_num == label_num
assert expected_code == label_code
assert expected_value == label_value
@@ -567,13 +514,13 @@ def test_check_label_name_2(self):
"""
Unit test check label name 2
"""
- xpl = SmartExplainer(self.model, label_dict = None)
+ xpl = SmartExplainer(self.model, label_dict=None)
xpl._classes = [1, 2]
entry = 1
expected_num = 0
expected_code = 1
expected_value = 1
- label_num, label_code, label_value = xpl.check_label_name(entry, 'code')
+ label_num, label_code, label_value = xpl.check_label_name(entry, "code")
assert expected_num == label_num
assert expected_code == label_code
assert expected_value == label_value
@@ -582,15 +529,15 @@ def test_check_label_name_3(self):
"""
Unit test check label name 3
"""
- label_dict = {1: 'Age', 2: 'Education'}
+ label_dict = {1: "Age", 2: "Education"}
xpl = SmartExplainer(self.model, label_dict=label_dict)
xpl.inv_label_dict = {v: k for k, v in xpl.label_dict.items()}
xpl._classes = [1, 2]
entry = 0
expected_num = 0
expected_code = 1
- expected_value = 'Age'
- label_num, label_code, label_value = xpl.check_label_name(entry, 'num')
+ expected_value = "Age"
+ label_num, label_code, label_value = xpl.check_label_name(entry, "num")
assert expected_num == label_num
assert expected_code == label_code
assert expected_value == label_value
@@ -601,34 +548,34 @@ def test_check_label_name_4(self):
"""
xpl = SmartExplainer(self.model)
label = 0
- origin = 'error'
+ origin = "error"
expected_msg = "Origin must be 'num', 'code' or 'value'."
- self.assertRaisesWithMessage(expected_msg, xpl.check_label_name, **{'label': label, 'origin': origin})
+ self.assertRaisesWithMessage(expected_msg, xpl.check_label_name, **{"label": label, "origin": origin})
def test_check_label_name_5(self):
"""
Unit test check label name 5
"""
- label_dict = {1: 'Age', 2: 'Education'}
+ label_dict = {1: "Age", 2: "Education"}
xpl = SmartExplainer(self.model, label_dict=label_dict)
xpl.inv_label_dict = {v: k for k, v in xpl.label_dict.items()}
xpl._classes = [1, 2]
- label = 'Absent'
+ label = "Absent"
expected_msg = f"Label (Absent) not found for origin (value)"
- origin = 'value'
- self.assertRaisesWithMessage(expected_msg, xpl.check_label_name, **{'label': label, 'origin': origin})
+ origin = "value"
+ self.assertRaisesWithMessage(expected_msg, xpl.check_label_name, **{"label": label, "origin": origin})
def test_check_features_name_1(self):
"""
Unit test check features name 1
"""
xpl = SmartExplainer(self.model)
- xpl.features_dict = {'tech_0': 'domain_0', 'tech_1': 'domain_1', 'tech_2': 'domain_2'}
+ xpl.features_dict = {"tech_0": "domain_0", "tech_1": "domain_1", "tech_2": "domain_2"}
xpl.inv_features_dict = {v: k for k, v in xpl.features_dict.items()}
- xpl.columns_dict = {0: 'tech_0', 1: 'tech_1', 2: 'tech_2'}
+ xpl.columns_dict = {0: "tech_0", 1: "tech_1", 2: "tech_2"}
xpl.inv_columns_dict = {v: k for k, v in xpl.columns_dict.items()}
- feature_list_1 = ['domain_0', 'tech_1']
- feature_list_2 = ['domain_0', 0]
+ feature_list_1 = ["domain_0", "tech_1"]
+ feature_list_2 = ["domain_0", 0]
self.assertRaises(ValueError, xpl.check_features_name, feature_list_1)
self.assertRaises(ValueError, xpl.check_features_name, feature_list_2)
@@ -637,11 +584,11 @@ def test_check_features_name_2(self):
Unit test check features name 2
"""
xpl = SmartExplainer(self.model)
- xpl.features_dict = {'tech_0': 'domain_0', 'tech_1': 'domain_1', 'tech_2': 'domain_2'}
+ xpl.features_dict = {"tech_0": "domain_0", "tech_1": "domain_1", "tech_2": "domain_2"}
xpl.inv_features_dict = {v: k for k, v in xpl.features_dict.items()}
- xpl.columns_dict = {0: 'tech_0', 1: 'tech_1', 2: 'tech_2'}
+ xpl.columns_dict = {0: "tech_0", 1: "tech_1", 2: "tech_2"}
xpl.inv_columns_dict = {v: k for k, v in xpl.columns_dict.items()}
- feature_list = ['domain_0', 'domain_2']
+ feature_list = ["domain_0", "domain_2"]
output = xpl.check_features_name(feature_list)
expected_output = [0, 2]
np.testing.assert_array_equal(output, expected_output)
@@ -651,9 +598,9 @@ def test_check_features_name_3(self):
Unit test check features name 3
"""
xpl = SmartExplainer(self.model)
- xpl.columns_dict = {0: 'tech_0', 1: 'tech_1', 2: 'tech_2'}
+ xpl.columns_dict = {0: "tech_0", 1: "tech_1", 2: "tech_2"}
xpl.inv_columns_dict = {v: k for k, v in xpl.columns_dict.items()}
- feature_list = ['tech_2']
+ feature_list = ["tech_2"]
output = xpl.check_features_name(feature_list)
expected_output = [2]
np.testing.assert_array_equal(output, expected_output)
@@ -686,14 +633,14 @@ def test_load_1(self):
temp, xpl = init_sme_to_pickle_test()
current = Path(path.abspath(__file__)).parent.parent.parent
- if str(sys.version)[0:3] == '3.8':
- pkl_file = path.join(current, 'data/xpl_to_load_38.pkl')
- elif str(sys.version)[0:3] == '3.9':
- pkl_file = path.join(current, 'data/xpl_to_load_39.pkl')
- elif str(sys.version)[0:4] == '3.10':
- pkl_file = path.join(current, 'data/xpl_to_load_310.pkl')
- elif str(sys.version)[0:4] == '3.11':
- pkl_file = path.join(current, 'data/xpl_to_load_311.pkl')
+ if str(sys.version)[0:3] == "3.8":
+ pkl_file = path.join(current, "data/xpl_to_load_38.pkl")
+ elif str(sys.version)[0:3] == "3.9":
+ pkl_file = path.join(current, "data/xpl_to_load_39.pkl")
+ elif str(sys.version)[0:4] == "3.10":
+ pkl_file = path.join(current, "data/xpl_to_load_310.pkl")
+ elif str(sys.version)[0:4] == "3.11":
+ pkl_file = path.join(current, "data/xpl_to_load_311.pkl")
else:
raise NotImplementedError
@@ -727,8 +674,8 @@ def test_predict_1(self):
Test predict method 1
"""
X = pd.DataFrame([[1, 1, 1], [2, 2, 2], [3, 3, 3]])
- y_true = pd.DataFrame(data=np.array([1, 2, 3]), columns=['pred'])
- y_false = pd.DataFrame(data=np.array([1, 2, 4]), columns=['pred'])
+ y_true = pd.DataFrame(data=np.array([1, 2, 3]), columns=["pred"])
+ y_false = pd.DataFrame(data=np.array([1, 2, 4]), columns=["pred"])
model = DecisionTreeRegressor().fit(X, y_true)
xpl = SmartExplainer(model)
xpl.compile(x=X, y_pred=y_false)
@@ -741,7 +688,7 @@ def test_predict_2(self):
Test predict method 2
"""
X = pd.DataFrame([[1, 1, 1], [2, 2, 2], [3, 3, 3]])
- y_true = pd.DataFrame(data=np.array([1, 2, 3]), columns=['pred'])
+ y_true = pd.DataFrame(data=np.array([1, 2, 3]), columns=["pred"])
model = DecisionTreeRegressor().fit(X, y_true)
xpl = SmartExplainer(model)
xpl.compile(x=X)
@@ -754,17 +701,17 @@ def test_predict_3(self):
Test predict method 3
"""
X = pd.DataFrame([[1, 1, 1], [2, 2, 2], [3, 3, 3]])
- y_target = pd.DataFrame(data=np.array([1, 2, 3]), columns=['pred'])
+ y_target = pd.DataFrame(data=np.array([1, 2, 3]), columns=["pred"])
model = DecisionTreeRegressor().fit(X, y_target)
xpl = SmartExplainer(model)
xpl.compile(x=X, y_target=y_target)
- xpl.predict() # prediction errors should be computed
+ xpl.predict() # prediction errors should be computed
assert xpl.prediction_error is not None
def test_add_1(self):
xpl = SmartExplainer(self.model)
- dataframe_yp = pd.Series([1, 3, 1], name='pred', index=[0, 1, 2])
+ dataframe_yp = pd.Series([1, 3, 1], name="pred", index=[0, 1, 2])
xpl.x_init = dataframe_yp
xpl.add(y_pred=dataframe_yp)
expected = SmartExplainer(self.model)
@@ -778,19 +725,19 @@ def test_add_2(self):
xpl = SmartExplainer(self.model)
xpl._classes = [0, 1]
xpl._case = "classification"
- xpl.add(label_dict={0: 'Zero', 1: 'One'})
- assert xpl.label_dict[0] == 'Zero'
- assert xpl.label_dict[1] == 'One'
+ xpl.add(label_dict={0: "Zero", 1: "One"})
+ assert xpl.label_dict[0] == "Zero"
+ assert xpl.label_dict[1] == "One"
def test_add_3(self):
"""
Unit test add 3
"""
xpl = SmartExplainer(self.model)
- xpl.columns_dict = {0: 'Age', 1: 'Education', 2: 'Sex'}
- xpl.add(features_dict={'Age': 'Age (Years Old)'})
- assert xpl.features_dict['Age'] == 'Age (Years Old)'
- assert xpl.features_dict['Education'] == 'Education'
+ xpl.columns_dict = {0: "Age", 1: "Education", 2: "Sex"}
+ xpl.add(features_dict={"Age": "Age (Years Old)"})
+ assert xpl.features_dict["Age"] == "Age (Years Old)"
+ assert xpl.features_dict["Education"] == "Education"
def test_to_pandas_1(self):
"""
@@ -799,81 +746,67 @@ def test_to_pandas_1(self):
xpl = SmartExplainer(self.model)
xpl.state = SmartState()
data = {}
- data['contrib_sorted'] = pd.DataFrame(
- [[0.32230754, 0.1550689, 0.10183475, 0.05471339],
- [-0.58547512, -0.37050409, -0.07249285, 0.00171975],
- [-0.48666675, 0.25507156, -0.16968889, 0.0757443]],
- columns=['contribution_0', 'contribution_1', 'contribution_2', 'contribution_3'],
- index=[0, 1, 2]
+ data["contrib_sorted"] = pd.DataFrame(
+ [
+ [0.32230754, 0.1550689, 0.10183475, 0.05471339],
+ [-0.58547512, -0.37050409, -0.07249285, 0.00171975],
+ [-0.48666675, 0.25507156, -0.16968889, 0.0757443],
+ ],
+ columns=["contribution_0", "contribution_1", "contribution_2", "contribution_3"],
+ index=[0, 1, 2],
)
- data['var_dict'] = pd.DataFrame(
- [[1, 0, 2, 3],
- [1, 0, 3, 2],
- [1, 0, 2, 3]],
- columns=['feature_0', 'feature_1', 'feature_2', 'feature_3'],
- index=[0, 1, 2]
+ data["var_dict"] = pd.DataFrame(
+ [[1, 0, 2, 3], [1, 0, 3, 2], [1, 0, 2, 3]],
+ columns=["feature_0", "feature_1", "feature_2", "feature_3"],
+ index=[0, 1, 2],
)
- data['x_sorted'] = pd.DataFrame(
- [[1., 3., 22., 1.],
- [2., 1., 2., 38.],
- [2., 3., 26., 1.]],
- columns=['feature_0', 'feature_1', 'feature_2', 'feature_3'],
- index=[0, 1, 2]
+ data["x_sorted"] = pd.DataFrame(
+ [[1.0, 3.0, 22.0, 1.0], [2.0, 1.0, 2.0, 38.0], [2.0, 3.0, 26.0, 1.0]],
+ columns=["feature_0", "feature_1", "feature_2", "feature_3"],
+ index=[0, 1, 2],
)
xpl.data = data
- xpl.columns_dict = {0: 'Pclass', 1: 'Sex', 2: 'Age', 3: 'Embarked'}
- xpl.features_dict = {'Pclass': 'Pclass', 'Sex': 'Sex', 'Age': 'Age', 'Embarked': 'Embarked'}
+ xpl.columns_dict = {0: "Pclass", 1: "Sex", 2: "Age", 3: "Embarked"}
+ xpl.features_dict = {"Pclass": "Pclass", "Sex": "Sex", "Age": "Age", "Embarked": "Embarked"}
xpl.x = pd.DataFrame(
- [[3., 1., 22., 1.],
- [1., 2., 38., 2.],
- [3., 2., 26., 1.]],
- columns=['Pclass', 'Sex', 'Age', 'Embarked'],
- index=[0, 1, 2]
+ [[3.0, 1.0, 22.0, 1.0], [1.0, 2.0, 38.0, 2.0], [3.0, 2.0, 26.0, 1.0]],
+ columns=["Pclass", "Sex", "Age", "Embarked"],
+ index=[0, 1, 2],
)
xpl.x_init = xpl.x
- xpl.contributions = data['contrib_sorted']
- xpl.y_pred = pd.DataFrame([1, 2, 3], columns=['pred'], index=[0, 1, 2])
- model = lambda : None
+ xpl.contributions = data["contrib_sorted"]
+ xpl.y_pred = pd.DataFrame([1, 2, 3], columns=["pred"], index=[0, 1, 2])
+ model = lambda: None
model.predict = types.MethodType(self.predict, model)
xpl.model = model
xpl._case, xpl._classes = check_model(model)
xpl.state = SmartState()
output = xpl.to_pandas(max_contrib=2)
expected = pd.DataFrame(
- [[1, 'Sex', 1.0, 0.32230754, 'Pclass', 3.0, 0.1550689],
- [2, 'Sex', 2.0, -0.58547512, 'Pclass', 1.0, -0.37050409],
- [3, 'Sex', 2.0, -0.48666675, 'Pclass', 3.0, 0.25507156]],
- columns=['pred',
- 'feature_1',
- 'value_1',
- 'contribution_1',
- 'feature_2',
- 'value_2',
- 'contribution_2'],
+ [
+ [1, "Sex", 1.0, 0.32230754, "Pclass", 3.0, 0.1550689],
+ [2, "Sex", 2.0, -0.58547512, "Pclass", 1.0, -0.37050409],
+ [3, "Sex", 2.0, -0.48666675, "Pclass", 3.0, 0.25507156],
+ ],
+ columns=["pred", "feature_1", "value_1", "contribution_1", "feature_2", "value_2", "contribution_2"],
index=[0, 1, 2],
- dtype=object
+ dtype=object,
)
- expected['pred'] = expected['pred'].astype(int)
+ expected["pred"] = expected["pred"].astype(int)
assert not pd.testing.assert_frame_equal(expected, output)
def predict_proba(self, arg1, arg2):
"""
predict_proba method
"""
- matrx = np.array(
- [[0.2, 0.8],
- [0.3, 0.7],
- [0.4, 0.6]]
- )
+ matrx = np.array([[0.2, 0.8], [0.3, 0.7], [0.4, 0.6]])
return matrx
def predict(self, arg1, arg2):
"""
predict method
"""
- matrx = np.array(
- [12, 3, 7]
- )
+ matrx = np.array([12, 3, 7])
return matrx
def test_to_pandas_2(self):
@@ -883,35 +816,45 @@ def test_to_pandas_2(self):
predict_proba output and column_dict attribute
"""
contrib = pd.DataFrame(
- [[0.32230754, 0.1550689, 0.10183475, 0.05471339],
- [-0.58547512, -0.37050409, -0.07249285, 0.00171975],
- [-0.48666675, 0.25507156, -0.16968889, 0.0757443]],
- index=[0, 1, 2]
- )
- x = pd.DataFrame(
- [[3., 1., 22., 1.],
- [1., 2., 38., 2.],
- [3., 2., 26., 1.]],
- index=[0, 1, 2]
+ [
+ [0.32230754, 0.1550689, 0.10183475, 0.05471339],
+ [-0.58547512, -0.37050409, -0.07249285, 0.00171975],
+ [-0.48666675, 0.25507156, -0.16968889, 0.0757443],
+ ],
+ index=[0, 1, 2],
)
- pred = pd.DataFrame([3, 1, 1], columns=['pred'], index=[0, 1, 2])
+ x = pd.DataFrame([[3.0, 1.0, 22.0, 1.0], [1.0, 2.0, 38.0, 2.0], [3.0, 2.0, 26.0, 1.0]], index=[0, 1, 2])
+ pred = pd.DataFrame([3, 1, 1], columns=["pred"], index=[0, 1, 2])
model = CatBoostClassifier().fit(x, pred)
xpl = SmartExplainer(model)
xpl.compile(contributions=contrib, x=x, y_pred=pred)
- xpl.columns_dict = {0: 'Pclass', 1: 'Sex', 2: 'Age', 3: 'Embarked'}
- xpl.features_dict = {'Pclass': 'Pclass', 'Sex': 'Sex', 'Age': 'Age', 'Embarked': 'Embarked'}
+ xpl.columns_dict = {0: "Pclass", 1: "Sex", 2: "Age", 3: "Embarked"}
+ xpl.features_dict = {"Pclass": "Pclass", "Sex": "Sex", "Age": "Age", "Embarked": "Embarked"}
output = xpl.to_pandas(max_contrib=3, positive=True, proba=True)
expected = pd.DataFrame(
- [[3, 0.8, 'Pclass', 3.0, 0.32230754, 'Sex', 1.0, 0.1550689, 'Age', 22.0, 0.10183475],
- [1, 0.3, 'Pclass', 1.0, 0.58547512, 'Sex', 2.0, 0.37050409, 'Age', 38.0, 0.07249285],
- [1, 0.4, 'Pclass', 3.0, 0.48666675, 'Age', 26.0, 0.16968889, np.nan, np.nan, np.nan]],
- columns=['pred', 'proba', 'feature_1', 'value_1', 'contribution_1', 'feature_2',
- 'value_2', 'contribution_2', 'feature_3', 'value_3', 'contribution_3'],
+ [
+ [3, 0.8, "Pclass", 3.0, 0.32230754, "Sex", 1.0, 0.1550689, "Age", 22.0, 0.10183475],
+ [1, 0.3, "Pclass", 1.0, 0.58547512, "Sex", 2.0, 0.37050409, "Age", 38.0, 0.07249285],
+ [1, 0.4, "Pclass", 3.0, 0.48666675, "Age", 26.0, 0.16968889, np.nan, np.nan, np.nan],
+ ],
+ columns=[
+ "pred",
+ "proba",
+ "feature_1",
+ "value_1",
+ "contribution_1",
+ "feature_2",
+ "value_2",
+ "contribution_2",
+ "feature_3",
+ "value_3",
+ "contribution_3",
+ ],
index=[0, 1, 2],
- dtype=object
+ dtype=object,
)
- expected['pred'] = expected['pred'].astype(int)
- expected['proba'] = expected['proba'].astype(float)
+ expected["pred"] = expected["pred"].astype(int)
+ expected["proba"] = expected["proba"].astype(float)
pd.testing.assert_series_equal(expected.dtypes, output.dtypes)
def test_to_pandas_3(self):
@@ -921,73 +864,63 @@ def test_to_pandas_3(self):
xpl = SmartExplainer(self.model)
xpl.state = SmartState()
data = {}
- data['contrib_sorted'] = pd.DataFrame(
- [[0.32230754, 0.1550689, 0.10183475, 0.05471339],
- [-0.58547512, -0.37050409, -0.07249285, 0.00171975],
- [-0.48666675, 0.25507156, -0.16968889, 0.0757443]],
- columns=['contribution_0', 'contribution_1', 'contribution_2', 'contribution_3'],
- index=[0, 1, 2]
+ data["contrib_sorted"] = pd.DataFrame(
+ [
+ [0.32230754, 0.1550689, 0.10183475, 0.05471339],
+ [-0.58547512, -0.37050409, -0.07249285, 0.00171975],
+ [-0.48666675, 0.25507156, -0.16968889, 0.0757443],
+ ],
+ columns=["contribution_0", "contribution_1", "contribution_2", "contribution_3"],
+ index=[0, 1, 2],
)
- data['var_dict'] = pd.DataFrame(
- [[1, 0, 2, 3],
- [1, 0, 3, 2],
- [1, 0, 2, 3]],
- columns=['feature_0', 'feature_1', 'feature_2', 'feature_3'],
- index=[0, 1, 2]
+ data["var_dict"] = pd.DataFrame(
+ [[1, 0, 2, 3], [1, 0, 3, 2], [1, 0, 2, 3]],
+ columns=["feature_0", "feature_1", "feature_2", "feature_3"],
+ index=[0, 1, 2],
)
- data['x_sorted'] = pd.DataFrame(
- [[1., 3., 22., 1.],
- [2., 1., 2., 38.],
- [2., 3., 26., 1.]],
- columns=['feature_0', 'feature_1', 'feature_2', 'feature_3'],
- index=[0, 1, 2]
+ data["x_sorted"] = pd.DataFrame(
+ [[1.0, 3.0, 22.0, 1.0], [2.0, 1.0, 2.0, 38.0], [2.0, 3.0, 26.0, 1.0]],
+ columns=["feature_0", "feature_1", "feature_2", "feature_3"],
+ index=[0, 1, 2],
)
data_groups = dict()
- data_groups['contrib_sorted'] = pd.DataFrame(
- [[0.32230754, 0.10183475, 0.05471339],
- [-0.58547512, -0.07249285, 0.00171975],
- [-0.48666675, -0.16968889, 0.0757443]],
- columns=['contribution_0', 'contribution_1', 'contribution_2'],
- index=[0, 1, 2]
+ data_groups["contrib_sorted"] = pd.DataFrame(
+ [
+ [0.32230754, 0.10183475, 0.05471339],
+ [-0.58547512, -0.07249285, 0.00171975],
+ [-0.48666675, -0.16968889, 0.0757443],
+ ],
+ columns=["contribution_0", "contribution_1", "contribution_2"],
+ index=[0, 1, 2],
)
- data_groups['var_dict'] = pd.DataFrame(
- [[0, 1, 2],
- [0, 2, 1],
- [0, 1, 2]],
- columns=['feature_0', 'feature_1', 'feature_2'],
- index=[0, 1, 2]
+ data_groups["var_dict"] = pd.DataFrame(
+ [[0, 1, 2], [0, 2, 1], [0, 1, 2]], columns=["feature_0", "feature_1", "feature_2"], index=[0, 1, 2]
)
- data_groups['x_sorted'] = pd.DataFrame(
- [[1., 22., 1.],
- [2., 2., 38.],
- [2., 26., 1.]],
- columns=['feature_0', 'feature_1', 'feature_2'],
- index=[0, 1, 2]
+ data_groups["x_sorted"] = pd.DataFrame(
+ [[1.0, 22.0, 1.0], [2.0, 2.0, 38.0], [2.0, 26.0, 1.0]],
+ columns=["feature_0", "feature_1", "feature_2"],
+ index=[0, 1, 2],
)
xpl.data = data
xpl.data_groups = data_groups
- xpl.columns_dict = {0: 'Pclass', 1: 'Sex', 2: 'Age', 3: 'Embarked'}
- xpl.features_dict = {'Pclass': 'Pclass', 'Sex': 'Sex', 'Age': 'Age', 'Embarked': 'Embarked', 'group1': 'group1'}
- xpl.features_groups = {'group1': ['Pclass', 'Sex']}
+ xpl.columns_dict = {0: "Pclass", 1: "Sex", 2: "Age", 3: "Embarked"}
+ xpl.features_dict = {"Pclass": "Pclass", "Sex": "Sex", "Age": "Age", "Embarked": "Embarked", "group1": "group1"}
+ xpl.features_groups = {"group1": ["Pclass", "Sex"]}
xpl.x = pd.DataFrame(
- [[3., 1., 22., 1.],
- [1., 2., 38., 2.],
- [3., 2., 26., 1.]],
- columns=['Pclass', 'Sex', 'Age', 'Embarked'],
- index=[0, 1, 2]
+ [[3.0, 1.0, 22.0, 1.0], [1.0, 2.0, 38.0, 2.0], [3.0, 2.0, 26.0, 1.0]],
+ columns=["Pclass", "Sex", "Age", "Embarked"],
+ index=[0, 1, 2],
)
xpl.x_init_groups = pd.DataFrame(
- [[3., 22., 1.],
- [1., 38., 2.],
- [3., 26., 1.]],
- columns=['group1', 'Age', 'Embarked'],
- index=[0, 1, 2]
+ [[3.0, 22.0, 1.0], [1.0, 38.0, 2.0], [3.0, 26.0, 1.0]],
+ columns=["group1", "Age", "Embarked"],
+ index=[0, 1, 2],
)
xpl.x_init = xpl.x
- xpl.contributions = data['contrib_sorted']
- xpl.y_pred = pd.DataFrame([1, 2, 3], columns=['pred'], index=[0, 1, 2])
+ xpl.contributions = data["contrib_sorted"]
+ xpl.y_pred = pd.DataFrame([1, 2, 3], columns=["pred"], index=[0, 1, 2])
model = lambda: None
model.predict = types.MethodType(self.predict, model)
xpl.model = model
@@ -996,20 +929,16 @@ def test_to_pandas_3(self):
output = xpl.to_pandas(max_contrib=2, use_groups=True)
expected = pd.DataFrame(
- [[1, 'group1', 1.0, 0.322308, 'Age', 22.0, 0.101835],
- [2, 'group1', 2.0, -0.585475, 'Embarked', 2.0, -0.072493],
- [3, 'group1', 2.0, -0.486667, 'Age', 26.0, -0.169689]],
- columns=['pred',
- 'feature_1',
- 'value_1',
- 'contribution_1',
- 'feature_2',
- 'value_2',
- 'contribution_2'],
+ [
+ [1, "group1", 1.0, 0.322308, "Age", 22.0, 0.101835],
+ [2, "group1", 2.0, -0.585475, "Embarked", 2.0, -0.072493],
+ [3, "group1", 2.0, -0.486667, "Age", 26.0, -0.169689],
+ ],
+ columns=["pred", "feature_1", "value_1", "contribution_1", "feature_2", "value_2", "contribution_2"],
index=[0, 1, 2],
- dtype=object
+ dtype=object,
)
- expected['pred'] = expected['pred'].astype(int)
+ expected["pred"] = expected["pred"].astype(int)
pd.testing.assert_frame_equal(expected, output)
def test_compute_features_import_1(self):
@@ -1019,18 +948,16 @@ def test_compute_features_import_1(self):
"""
xpl = SmartExplainer(self.model)
contributions = pd.DataFrame(
- [[1, 2, 3, 4],
- [5, 6, 7, 8],
- [9, 10, 11, 12]],
- columns=['contribution_0', 'contribution_1', 'contribution_2', 'contribution_3'],
- index=[0, 1, 2]
+ [[1, 2, 3, 4], [5, 6, 7, 8], [9, 10, 11, 12]],
+ columns=["contribution_0", "contribution_1", "contribution_2", "contribution_3"],
+ index=[0, 1, 2],
)
xpl.features_imp = None
xpl.contributions = contributions
xpl.backend = ShapBackend(model=DecisionTreeClassifier().fit([[0]], [[0]]))
xpl.backend.state = SmartState()
xpl.explain_data = None
- xpl._case = 'regression'
+ xpl._case = "regression"
xpl.compute_features_import()
expected = contributions.abs().sum().sort_values(ascending=True)
expected = expected / expected.sum()
@@ -1043,18 +970,14 @@ def test_compute_features_import_2(self):
"""
xpl = SmartExplainer(self.model)
contrib1 = pd.DataFrame(
- [[1, 2, 3, 4],
- [5, 6, 7, 8],
- [9, 10, 11, 12]],
- columns=['contribution_0', 'contribution_1', 'contribution_2', 'contribution_3'],
- index=[0, 1, 2]
+ [[1, 2, 3, 4], [5, 6, 7, 8], [9, 10, 11, 12]],
+ columns=["contribution_0", "contribution_1", "contribution_2", "contribution_3"],
+ index=[0, 1, 2],
)
contrib2 = pd.DataFrame(
- [[13, 14, 15, 16],
- [17, 18, 19, 20],
- [21, 22, 23, 24]],
- columns=['contribution_0', 'contribution_1', 'contribution_2', 'contribution_3'],
- index=[0, 1, 2]
+ [[13, 14, 15, 16], [17, 18, 19, 20], [21, 22, 23, 24]],
+ columns=["contribution_0", "contribution_1", "contribution_2", "contribution_3"],
+ index=[0, 1, 2],
)
contributions = [contrib1, contrib2]
xpl.features_imp = None
@@ -1075,47 +998,43 @@ def test_to_smartpredictor_1(self):
"""
Unit test 1 to_smartpredictor
"""
- df = pd.DataFrame(range(0, 5), columns=['id'])
- df['y'] = df['id'].apply(lambda x: 1 if x < 2 else 0)
- df['x1'] = np.random.randint(1, 123, df.shape[0])
- df['x2'] = ["S", "M", "S", "D", "M"]
- df = df.set_index('id')
+ df = pd.DataFrame(range(0, 5), columns=["id"])
+ df["y"] = df["id"].apply(lambda x: 1 if x < 2 else 0)
+ df["x1"] = np.random.randint(1, 123, df.shape[0])
+ df["x2"] = ["S", "M", "S", "D", "M"]
+ df = df.set_index("id")
encoder = ce.OrdinalEncoder(cols=["x2"], handle_unknown="None")
encoder_fitted = encoder.fit(df[["x1", "x2"]])
df_encoded = encoder_fitted.transform(df[["x1", "x2"]])
- clf = cb.CatBoostClassifier(n_estimators=1).fit(df_encoded[['x1', 'x2']], df['y'])
-
- postprocessing = {"x2": {
- "type": "transcoding",
- "rule": {"S": "single", "M": "married", "D": "divorced"}}}
- xpl = SmartExplainer(clf, features_dict={"x1": "age", "x2": "family_situation"},
- preprocessing=encoder_fitted,
- postprocessing=postprocessing)
+ clf = cb.CatBoostClassifier(n_estimators=1).fit(df_encoded[["x1", "x2"]], df["y"])
+
+ postprocessing = {"x2": {"type": "transcoding", "rule": {"S": "single", "M": "married", "D": "divorced"}}}
+ xpl = SmartExplainer(
+ clf,
+ features_dict={"x1": "age", "x2": "family_situation"},
+ preprocessing=encoder_fitted,
+ postprocessing=postprocessing,
+ )
- xpl.compile(x=df_encoded[['x1', 'x2']])
+ xpl.compile(x=df_encoded[["x1", "x2"]])
predictor_1 = xpl.to_smartpredictor()
- xpl.mask_params = {
- 'features_to_hide': None,
- 'threshold': None,
- 'positive': True,
- 'max_contrib': 1
- }
+ xpl.mask_params = {"features_to_hide": None, "threshold": None, "positive": True, "max_contrib": 1}
predictor_2 = xpl.to_smartpredictor()
- assert hasattr(predictor_1, 'model')
- assert hasattr(predictor_1, 'backend')
- assert hasattr(predictor_1, 'features_dict')
- assert hasattr(predictor_1, 'label_dict')
- assert hasattr(predictor_1, '_case')
- assert hasattr(predictor_1, '_classes')
- assert hasattr(predictor_1, 'columns_dict')
- assert hasattr(predictor_1, 'features_types')
- assert hasattr(predictor_1, 'preprocessing')
- assert hasattr(predictor_1, 'postprocessing')
- assert hasattr(predictor_1, 'mask_params')
- assert hasattr(predictor_2, 'mask_params')
+ assert hasattr(predictor_1, "model")
+ assert hasattr(predictor_1, "backend")
+ assert hasattr(predictor_1, "features_dict")
+ assert hasattr(predictor_1, "label_dict")
+ assert hasattr(predictor_1, "_case")
+ assert hasattr(predictor_1, "_classes")
+ assert hasattr(predictor_1, "columns_dict")
+ assert hasattr(predictor_1, "features_types")
+ assert hasattr(predictor_1, "preprocessing")
+ assert hasattr(predictor_1, "postprocessing")
+ assert hasattr(predictor_1, "mask_params")
+ assert hasattr(predictor_2, "mask_params")
assert predictor_1.model == xpl.model
assert predictor_1.backend == xpl.backend
@@ -1126,25 +1045,27 @@ def test_to_smartpredictor_1(self):
assert predictor_1.columns_dict == xpl.columns_dict
assert predictor_1.preprocessing == xpl.preprocessing
assert predictor_1.postprocessing == xpl.postprocessing
- assert all(predictor_1.features_types[feature] == str(xpl.x_init[feature].dtypes)
- for feature in xpl.x_init.columns)
+ assert all(
+ predictor_1.features_types[feature] == str(xpl.x_init[feature].dtypes) for feature in xpl.x_init.columns
+ )
assert predictor_2.mask_params == xpl.mask_params
def test_get_interaction_values_1(self):
- df = pd.DataFrame({
- "y": np.random.randint(2, size=10),
- "a": np.random.rand(10)*10,
- "b": np.random.rand(10),
- })
+ df = pd.DataFrame(
+ {
+ "y": np.random.randint(2, size=10),
+ "a": np.random.rand(10) * 10,
+ "b": np.random.rand(10),
+ }
+ )
- clf = RandomForestClassifier(n_estimators=5).fit(df[['a', 'b']], df['y'])
+ clf = RandomForestClassifier(n_estimators=5).fit(df[["a", "b"]], df["y"])
- xpl = SmartExplainer(clf,
- backend='shap',
- explainer_args={'explainer': shap.explainers.TreeExplainer,
- 'model': clf})
- xpl.compile(x=df.drop('y', axis=1))
+ xpl = SmartExplainer(
+ clf, backend="shap", explainer_args={"explainer": shap.explainers.TreeExplainer, "model": clf}
+ )
+ xpl.compile(x=df.drop("y", axis=1))
shap_interaction_values = xpl.get_interaction_values(n_samples_max=10)
assert shap_interaction_values.shape[0] == 10
@@ -1152,17 +1073,16 @@ def test_get_interaction_values_1(self):
shap_interaction_values = xpl.get_interaction_values()
assert shap_interaction_values.shape[0] == df.shape[0]
-
- @patch('shapash.explainer.smart_explainer.SmartApp')
- @patch('shapash.explainer.smart_explainer.CustomThread')
- @patch('shapash.explainer.smart_explainer.get_host_name')
+ @patch("shapash.explainer.smart_explainer.SmartApp")
+ @patch("shapash.explainer.smart_explainer.CustomThread")
+ @patch("shapash.explainer.smart_explainer.get_host_name")
def test_run_app_1(self, mock_get_host_name, mock_custom_thread, mock_smartapp):
"""
Test that when y_pred is not given, y_pred is automatically computed.
"""
X = pd.DataFrame([[1, 1, 1], [2, 2, 2], [3, 3, 3]])
contributions = pd.DataFrame([[0.1, -0.2, 0.3], [0.1, -0.2, 0.3], [0.1, -0.2, 0.3]])
- y_true = pd.DataFrame(data=np.array([1, 2, 3]), columns=['pred'])
+ y_true = pd.DataFrame(data=np.array([1, 2, 3]), columns=["pred"])
model = DecisionTreeRegressor().fit(X, y_true)
xpl = SmartExplainer(model)
xpl.compile(contributions=contributions, x=X)
@@ -1170,42 +1090,42 @@ def test_run_app_1(self, mock_get_host_name, mock_custom_thread, mock_smartapp):
assert xpl.y_pred is not None
- @patch('shapash.explainer.smart_explainer.SmartApp')
- @patch('shapash.explainer.smart_explainer.CustomThread')
- @patch('shapash.explainer.smart_explainer.get_host_name')
+ @patch("shapash.explainer.smart_explainer.SmartApp")
+ @patch("shapash.explainer.smart_explainer.CustomThread")
+ @patch("shapash.explainer.smart_explainer.get_host_name")
def test_run_app_2(self, mock_get_host_name, mock_custom_thread, mock_smartapp):
"""
Test that when y_target is given
"""
X = pd.DataFrame([[1, 1, 1], [2, 2, 2], [3, 3, 3]])
contributions = pd.DataFrame([[0.1, -0.2, 0.3], [0.1, -0.2, 0.3], [0.1, -0.2, 0.3]])
- y_true = pd.DataFrame(data=np.array([1, 2, 3]), columns=['pred'])
+ y_true = pd.DataFrame(data=np.array([1, 2, 3]), columns=["pred"])
model = DecisionTreeRegressor().fit(X, y_true)
xpl = SmartExplainer(model)
xpl.compile(contributions=contributions, x=X, y_target=y_true)
xpl.run_app()
assert xpl.y_target is not None
- @patch('shapash.report.generation.export_and_save_report')
- @patch('shapash.report.generation.execute_report')
+ @patch("shapash.report.generation.export_and_save_report")
+ @patch("shapash.report.generation.execute_report")
def test_generate_report(self, mock_execute_report, mock_export_and_save_report):
"""
Test generate report method
"""
- df = pd.DataFrame(range(0, 21), columns=['id'])
- df['y'] = df['id'].apply(lambda x: 1 if x < 10 else 0)
- df['x1'] = np.random.randint(1, 123, df.shape[0])
- df['x2'] = np.random.randint(1, 3, df.shape[0])
- df = df.set_index('id')
- clf = cb.CatBoostClassifier(n_estimators=1).fit(df[['x1', 'x2']], df['y'])
+ df = pd.DataFrame(range(0, 21), columns=["id"])
+ df["y"] = df["id"].apply(lambda x: 1 if x < 10 else 0)
+ df["x1"] = np.random.randint(1, 123, df.shape[0])
+ df["x2"] = np.random.randint(1, 3, df.shape[0])
+ df = df.set_index("id")
+ clf = cb.CatBoostClassifier(n_estimators=1).fit(df[["x1", "x2"]], df["y"])
xpl = SmartExplainer(clf)
- xpl.compile(x=df[['x1', 'x2']])
- xpl.generate_report(output_file='test', project_info_file='test')
+ xpl.compile(x=df[["x1", "x2"]])
+ xpl.generate_report(output_file="test", project_info_file="test")
mock_execute_report.assert_called_once()
mock_export_and_save_report.assert_called_once()
def test_compute_features_stability_1(self):
- df = pd.DataFrame(np.random.randint(1, 100, size=(15, 4)), columns=list('ABCD'))
+ df = pd.DataFrame(np.random.randint(1, 100, size=(15, 4)), columns=list("ABCD"))
selection = [1, 3]
X = df.iloc[:, :-1]
y = df.iloc[:, -1]
@@ -1221,7 +1141,7 @@ def test_compute_features_stability_1(self):
assert xpl.features_stability["amplitude"].shape == expected
def test_compute_features_stability_2(self):
- df = pd.DataFrame(np.random.randint(1, 100, size=(15, 4)), columns=list('ABCD'))
+ df = pd.DataFrame(np.random.randint(1, 100, size=(15, 4)), columns=list("ABCD"))
selection = [1]
X = df.iloc[:, :-1]
y = df.iloc[:, -1]
@@ -1236,7 +1156,7 @@ def test_compute_features_stability_2(self):
assert xpl.local_neighbors["norm_shap"].shape[1] == expected
def test_compute_features_compacity(self):
- df = pd.DataFrame(np.random.randint(0, 100, size=(15, 4)), columns=list('ABCD'))
+ df = pd.DataFrame(np.random.randint(0, 100, size=(15, 4)), columns=list("ABCD"))
selection = [1, 3]
X = df.iloc[:, :-1]
y = df.iloc[:, -1]
diff --git a/tests/unit_tests/explainer/test_smart_plotter.py b/tests/unit_tests/explainer/test_smart_plotter.py
index 4b0652f1..57f279a0 100644
--- a/tests/unit_tests/explainer/test_smart_plotter.py
+++ b/tests/unit_tests/explainer/test_smart_plotter.py
@@ -5,19 +5,21 @@
import types
import unittest
from unittest.mock import patch
-import pandas as pd
+
+import category_encoders as ce
import numpy as np
-import plotly.graph_objects as go
+import pandas as pd
import plotly.express as px
-from sklearn.tree import DecisionTreeRegressor, DecisionTreeClassifier
+import plotly.graph_objects as go
from catboost import CatBoostClassifier
-import category_encoders as ce
+from sklearn.tree import DecisionTreeClassifier, DecisionTreeRegressor
+
from shapash import SmartExplainer
from shapash.backend import ShapBackend
-from shapash.utils.check import check_model
from shapash.explainer.multi_decorator import MultiDecorator
from shapash.explainer.smart_state import SmartState
from shapash.style.style_utils import get_palette
+from shapash.utils.check import check_model
class TestSmartPlotter(unittest.TestCase):
@@ -29,6 +31,7 @@ class TestSmartPlotter(unittest.TestCase):
unittest : [type]
[description]
"""
+
def predict_proba(self, arg1, X):
"""
predict_proba method
@@ -44,9 +47,7 @@ def predict(self, arg1, arg2):
"""
predict method
"""
- matrx = np.array(
- [1, 3]
- )
+ matrx = np.array([1, 3])
return matrx
def setUp(self):
@@ -54,75 +55,50 @@ def setUp(self):
SetUp
"""
self.x_init = pd.DataFrame(
- data=np.array(
- [['PhD', 34],
- ['Master', 27]]),
- columns=['X1', 'X2'],
- index=['person_A', 'person_B']
+ data=np.array([["PhD", 34], ["Master", 27]]), columns=["X1", "X2"], index=["person_A", "person_B"]
)
self.contrib0 = pd.DataFrame(
- data=np.array(
- [[-3.4, 0.78],
- [1.2, 3.6]]),
- columns=['X1', 'X2'],
- index=['person_A', 'person_B']
+ data=np.array([[-3.4, 0.78], [1.2, 3.6]]), columns=["X1", "X2"], index=["person_A", "person_B"]
)
self.contrib1 = pd.DataFrame(
- data=np.array(
- [[-0.3, 0.89],
- [4.7, 0.6]]),
- columns=['X1', 'X2'],
- index=['person_A', 'person_B']
+ data=np.array([[-0.3, 0.89], [4.7, 0.6]]), columns=["X1", "X2"], index=["person_A", "person_B"]
)
self.contrib_sorted = pd.DataFrame(
- data=np.array(
- [[-3.4, 0.78],
- [3.6, 1.2]]),
- columns=['contrib_0', 'contrib_1'],
- index=['person_A', 'person_B']
+ data=np.array([[-3.4, 0.78], [3.6, 1.2]]),
+ columns=["contrib_0", "contrib_1"],
+ index=["person_A", "person_B"],
)
self.x_sorted = pd.DataFrame(
- data=np.array(
- [['PhD', 34],
- [27, 'Master']]),
- columns=['feature_0', 'feature_1'],
- index=['person_A', 'person_B']
+ data=np.array([["PhD", 34], [27, "Master"]]),
+ columns=["feature_0", "feature_1"],
+ index=["person_A", "person_B"],
)
self.var_dict = pd.DataFrame(
- data=np.array(
- [[0, 1],
- [1, 0]]),
- columns=['feature_0', 'feature_1'],
- index=['person_A', 'person_B']
+ data=np.array([[0, 1], [1, 0]]), columns=["feature_0", "feature_1"], index=["person_A", "person_B"]
)
self.mask = pd.DataFrame(
- data=np.array(
- [[True, False],
- [True, True]]),
- columns=['feature_0', 'feature_1'],
- index=['person_A', 'person_B']
- )
- self.features_compacity = {
- "features_needed": [1, 1],
- "distance_reached": np.array([0.12, 0.16])
- }
+ data=np.array([[True, False], [True, True]]),
+ columns=["feature_0", "feature_1"],
+ index=["person_A", "person_B"],
+ )
+ self.features_compacity = {"features_needed": [1, 1], "distance_reached": np.array([0.12, 0.16])}
encoder = ce.OrdinalEncoder(cols=["X1"], handle_unknown="None").fit(self.x_init)
model = CatBoostClassifier().fit(encoder.transform(self.x_init), [0, 1])
self.model = model
# Declare explainer object
- self.feature_dictionary = {'X1': 'Education', 'X2': 'Age'}
+ self.feature_dictionary = {"X1": "Education", "X2": "Age"}
self.smart_explainer = SmartExplainer(model, features_dict=self.feature_dictionary, preprocessing=encoder)
self.smart_explainer.data = dict()
- self.smart_explainer.data['contrib_sorted'] = self.contrib_sorted
- self.smart_explainer.data['x_sorted'] = self.x_sorted
- self.smart_explainer.data['var_dict'] = self.var_dict
+ self.smart_explainer.data["contrib_sorted"] = self.contrib_sorted
+ self.smart_explainer.data["x_sorted"] = self.x_sorted
+ self.smart_explainer.data["var_dict"] = self.var_dict
self.smart_explainer.x_encoded = encoder.transform(self.x_init)
self.smart_explainer.x_init = self.x_init
self.smart_explainer.postprocessing_modifications = False
self.smart_explainer.backend = ShapBackend(model=model)
self.smart_explainer.backend.state = MultiDecorator(SmartState())
self.smart_explainer.explain_data = None
- #self.smart_explainer.x_contrib_plot = self.x_contrib_plot
+ # self.smart_explainer.x_contrib_plot = self.x_contrib_plot
self.smart_explainer.columns_dict = {i: col for i, col in enumerate(self.smart_explainer.x_init.columns)}
self.smart_explainer.inv_columns_dict = {v: k for k, v in self.smart_explainer.columns_dict.items()}
self.smart_explainer.mask = self.mask
@@ -139,14 +115,14 @@ def test_define_style_attributes(self):
# clear style attributes
del self.smart_explainer.plot._style_dict
- colors_dict = get_palette('default')
+ colors_dict = get_palette("default")
self.smart_explainer.plot.define_style_attributes(colors_dict=colors_dict)
- assert hasattr(self.smart_explainer.plot, '_style_dict')
- assert len(list(self.smart_explainer.plot._style_dict.keys())) > 0
+ assert hasattr(self.smart_explainer.plot, "_style_dict")
+ assert len(list(self.smart_explainer.plot._style_dict.keys())) > 0
- @patch('shapash.explainer.smart_explainer.SmartExplainer.filter')
- @patch('shapash.explainer.smart_plotter.SmartPlotter.local_pred')
+ @patch("shapash.explainer.smart_explainer.SmartExplainer.filter")
+ @patch("shapash.explainer.smart_plotter.SmartPlotter.local_pred")
def test_local_plot_1(self, local_pred, filter):
"""
Unit test Local plot 1
@@ -158,33 +134,22 @@ def test_local_plot_1(self, local_pred, filter):
local_pred.return_value = 12.88
filter.return_value = None
self.smart_explainer._case = "regression"
- output = self.smart_explainer.plot.local_plot(index='person_B')
+ output = self.smart_explainer.plot.local_plot(index="person_B")
output_data = output.data
- feature_values = ['Age : Education : Age : Education : Age : Education : Age : Education : B Response: 1 - Proba: 0.5800 "
assert output.layout.title.text == tit
- @patch('shapash.explainer.smart_explainer.SmartExplainer.filter')
- @patch('shapash.explainer.smart_plotter.select_lines')
- @patch('shapash.explainer.smart_plotter.SmartPlotter.local_pred')
+ @patch("shapash.explainer.smart_explainer.SmartExplainer.filter")
+ @patch("shapash.explainer.smart_plotter.select_lines")
+ @patch("shapash.explainer.smart_plotter.SmartPlotter.local_pred")
def test_local_plot_5(self, local_pred, select_lines, filter):
"""
Unit test local plot 5
@@ -358,128 +262,66 @@ def test_local_plot_5(self, local_pred, select_lines, filter):
[description]
"""
local_pred.return_value = 0.58
- select_lines.return_value = ['B']
+ select_lines.return_value = ["B"]
filter.return_value = None
- index = ['A', 'B']
- x_init = pd.DataFrame(
- data=np.array(
- [['PhD', 34],
- ['Master', 27]]
- ),
- columns=['X1', 'X2'],
- index=index
- )
+ index = ["A", "B"]
+ x_init = pd.DataFrame(data=np.array([["PhD", 34], ["Master", 27]]), columns=["X1", "X2"], index=index)
contrib_sorted1 = pd.DataFrame(
- data=np.array(
- [[-3.4, 0.78],
- [3.6, 1.2]]
- ),
- columns=['contrib_0', 'contrib_1'],
- index=index
+ data=np.array([[-3.4, 0.78], [3.6, 1.2]]), columns=["contrib_0", "contrib_1"], index=index
)
contrib_sorted2 = pd.DataFrame(
- data=np.array(
- [[-0.4, 0.78],
- [0.6, 0.2]]
- ),
- columns=['contrib_0', 'contrib_1'],
- index=index
+ data=np.array([[-0.4, 0.78], [0.6, 0.2]]), columns=["contrib_0", "contrib_1"], index=index
)
x_sorted1 = pd.DataFrame(
- data=np.array(
- [['PhD', 34],
- [27, 'Master']]
- ),
- columns=['feature_0', 'feature_1'],
- index=index
+ data=np.array([["PhD", 34], [27, "Master"]]), columns=["feature_0", "feature_1"], index=index
)
x_sorted2 = pd.DataFrame(
- data=np.array(
- [['PhD', 34],
- [27, 'Master']]
- ),
- columns=['feature_0', 'feature_1'],
- index=index
- )
- var_dict1 = pd.DataFrame(
- data=np.array(
- [[0, 1],
- [1, 0]]
- ),
- columns=['feature_0', 'feature_1'],
- index=index
- )
- var_dict2 = pd.DataFrame(
- data=np.array(
- [[0, 1],
- [1, 0]]
- ),
- columns=['feature_0', 'feature_1'],
- index=index
+ data=np.array([["PhD", 34], [27, "Master"]]), columns=["feature_0", "feature_1"], index=index
)
+ var_dict1 = pd.DataFrame(data=np.array([[0, 1], [1, 0]]), columns=["feature_0", "feature_1"], index=index)
+ var_dict2 = pd.DataFrame(data=np.array([[0, 1], [1, 0]]), columns=["feature_0", "feature_1"], index=index)
mask1 = pd.DataFrame(
- data=np.array(
- [[True, False],
- [True, True]]
- ),
- columns=['feature_0', 'feature_1'],
- index=index)
+ data=np.array([[True, False], [True, True]]), columns=["feature_0", "feature_1"], index=index
+ )
mask2 = pd.DataFrame(
- data=np.array(
- [[False, True],
- [False, True]]
- ),
- columns=['feature_0', 'feature_1'],
- index=index)
+ data=np.array([[False, True], [False, True]]), columns=["feature_0", "feature_1"], index=index
+ )
mask_contrib1 = pd.DataFrame(
- data=np.array(
- [[0.0, 0.78],
- [0.0, 1.20]]
- ),
- columns=['masked_neg', 'masked_pos'],
- index=index)
+ data=np.array([[0.0, 0.78], [0.0, 1.20]]), columns=["masked_neg", "masked_pos"], index=index
+ )
mask_contrib2 = pd.DataFrame(
- data=np.array(
- [[0.0, 0.78],
- [0.0, 0.20]]
- ),
- columns=['masked_neg', 'masked_pos'],
- index=index)
-
- feature_dictionary = {'X1': 'Education', 'X2': 'Age'}
+ data=np.array([[0.0, 0.78], [0.0, 0.20]]), columns=["masked_neg", "masked_pos"], index=index
+ )
+
+ feature_dictionary = {"X1": "Education", "X2": "Age"}
smart_explainer_mi = SmartExplainer(model=self.model, features_dict=feature_dictionary)
smart_explainer_mi.data = dict()
smart_explainer_mi.contributions = [contrib_sorted1, contrib_sorted2]
- smart_explainer_mi.data['contrib_sorted'] = [contrib_sorted1, contrib_sorted2]
- smart_explainer_mi.data['x_sorted'] = [x_sorted1, x_sorted2]
- smart_explainer_mi.data['var_dict'] = [var_dict1, var_dict2]
+ smart_explainer_mi.data["contrib_sorted"] = [contrib_sorted1, contrib_sorted2]
+ smart_explainer_mi.data["x_sorted"] = [x_sorted1, x_sorted2]
+ smart_explainer_mi.data["var_dict"] = [var_dict1, var_dict2]
smart_explainer_mi.x_init = x_init
smart_explainer_mi.columns_dict = {i: col for i, col in enumerate(smart_explainer_mi.x_init.columns)}
smart_explainer_mi.mask = [mask1, mask2]
smart_explainer_mi.masked_contributions = [mask_contrib1, mask_contrib2]
- smart_explainer_mi.mask_params = {'features_to_hide': None,
- 'threshold': None,
- 'positive': None,
- 'max_contrib': 1}
+ smart_explainer_mi.mask_params = {
+ "features_to_hide": None,
+ "threshold": None,
+ "positive": None,
+ "max_contrib": 1,
+ }
smart_explainer_mi._case = "classification"
smart_explainer_mi._classes = [0, 1]
smart_explainer_mi.state = MultiDecorator(SmartState())
condition = "index == 'B'"
output = smart_explainer_mi.plot.local_plot(query=condition)
- feature_values = ['Hidden Positive Contributions ',
- 'Education : Hidden Positive Contributions ", "Education : Age : Education : Age : Education : {} : {} : X1 : X2 : Hidden Positive Contributions ']
+ ylabel = ["X1 : X2 : Hidden Positive Contributions "]
self.smart_explainer.masked_contributions = pd.DataFrame()
bars = []
comblist = list(zip(order, contributions, ylabel))
comblist.sort(reverse=True)
for elem in comblist:
- bars.append(go.Bar(
- x=[elem[1]],
- y=[elem[2]],
- orientation='h'
- ))
- expected_output_fig = go.Figure(
- data=bars,
- layout=go.Layout(yaxis=dict(type='category')))
+ bars.append(go.Bar(x=[elem[1]], y=[elem[2]], orientation="h"))
+ expected_output_fig = go.Figure(data=bars, layout=go.Layout(yaxis=dict(type="category")))
self.smart_explainer._case = "regression"
- fig_output = self.smart_explainer.plot.plot_bar_chart('ind', var_dict, x_val, contributions)
+ fig_output = self.smart_explainer.plot.plot_bar_chart("ind", var_dict, x_val, contributions)
for part in list(zip(fig_output.data, expected_output_fig.data)):
assert part[0].x == part[1].x
assert part[0].y == part[1].y
@@ -890,12 +686,14 @@ def test_contribution_plot_1(self):
"""
Classification
"""
- col = 'X1'
+ col = "X1"
output = self.smart_explainer.plot.contribution_plot(col, violin_maxf=0, proba=False)
- expected_output = go.Scatter(x=self.smart_explainer.x_init[col],
- y=self.smart_explainer.contributions[-1][col],
- mode='markers',
- hovertext=[f"Id: {x}Education - Feature ContributionResponse: 1 - Length of random Subset: 39 (98%) "
- assert output.layout.title['text'] == expected_title
+ assert output.layout.title["text"] == expected_title
def test_contribution_plot_10(self):
"""
Regression with pred and subset
"""
- col = 'X2'
+ col = "X2"
xpl = self.smart_explainer
xpl.x_init = pd.concat([xpl.x_init] * 4, ignore_index=True)
xpl.x_init.index = [i for i in range(xpl.x_init.shape[0])]
@@ -1115,15 +923,18 @@ def test_contribution_plot_10(self):
xpl.contributions = pd.concat([self.contrib1] * 4, ignore_index=True)
xpl._case = "regression"
xpl.state = SmartState()
- xpl.y_pred = pd.DataFrame([0.46989877093, 12.749302948]*4, columns=['pred'], index=xpl.x_init.index)
+ xpl.y_pred = pd.DataFrame([0.46989877093, 12.749302948] * 4, columns=["pred"], index=xpl.x_init.index)
subset = [1, 2, 6, 7]
output = xpl.plot.contribution_plot(col, selection=subset, violin_maxf=0)
- expected_output = go.Scatter(x=xpl.x_init[col].loc[subset],
- y=xpl.contributions[col].loc[subset],
- mode='markers',
- hovertext=[f"Id: {x}Age - Feature ContributionLength of user-defined Subset: 4 (50%) "
- assert output.layout.title['text'] == expected_title
+ assert output.layout.title["text"] == expected_title
def test_contribution_plot_11(self):
"""
classification with proba
"""
- col = 'X1'
+ col = "X1"
xpl = self.smart_explainer
xpl.proba_values = pd.DataFrame(
- data=np.array(
- [[0.4, 0.6],
- [0.3, 0.7]]),
- columns=['class_1', 'class_2'],
- index=xpl.x_encoded.index.values)
+ data=np.array([[0.4, 0.6], [0.3, 0.7]]), columns=["class_1", "class_2"], index=xpl.x_encoded.index.values
+ )
output = self.smart_explainer.plot.contribution_plot(col)
assert str(type(output.data[-1])) == ""
- self.assertListEqual(list(output.data[-1]['marker']['color']), [0.6, 0.7])
- self.assertListEqual(list(output.data[-1]['y']), [-0.3, 4.7])
+ self.assertListEqual(list(output.data[-1]["marker"]["color"]), [0.6, 0.7])
+ self.assertListEqual(list(output.data[-1]["y"]), [-0.3, 4.7])
def test_contribution_plot_12(self):
"""
contribution plot with groups of features for classification case
"""
x_encoded = pd.DataFrame(
- data=np.array(
- [[0, 34],
- [1, 27]]),
- columns=['X1', 'X2'],
- index=['person_A', 'person_B']
+ data=np.array([[0, 34], [1, 27]]), columns=["X1", "X2"], index=["person_A", "person_B"]
)
xpl = self.smart_explainer
xpl.inv_features_dict = {}
- col = 'group1'
+ col = "group1"
xpl.x_encoded = x_encoded
xpl.contributions[0] = pd.concat([xpl.contributions[0]] * 10, ignore_index=True)
xpl.contributions[1] = pd.concat([xpl.contributions[1]] * 10, ignore_index=True)
@@ -1172,7 +976,7 @@ def test_contribution_plot_12(self):
xpl.postprocessing_modifications = False
xpl.preprocessing = None
# Creates a group of features named group1
- xpl.features_groups = {'group1': ['X1', 'X2']}
+ xpl.features_groups = {"group1": ["X1", "X2"]}
xpl.contributions_groups = xpl.state.compute_grouped_contributions(xpl.contributions, xpl.features_groups)
xpl.features_imp_groups = None
xpl._update_features_dict_with_groups(features_groups=xpl.features_groups)
@@ -1180,7 +984,7 @@ def test_contribution_plot_12(self):
output = xpl.plot.contribution_plot(col, proba=False)
assert len(output.data) == 1
- assert output.data[0].type == 'scatter'
+ assert output.data[0].type == "scatter"
self.setUp()
def test_contribution_plot_13(self):
@@ -1188,15 +992,11 @@ def test_contribution_plot_13(self):
contribution plot with groups of features for regression case
"""
x_encoded = pd.DataFrame(
- data=np.array(
- [[0, 34],
- [1, 27]]),
- columns=['X1', 'X2'],
- index=['person_A', 'person_B']
+ data=np.array([[0, 34], [1, 27]]), columns=["X1", "X2"], index=["person_A", "person_B"]
)
xpl = self.smart_explainer
xpl.inv_features_dict = {}
- col = 'group1'
+ col = "group1"
xpl.x_encoded = x_encoded
xpl.contributions = pd.concat([self.contrib1] * 10, ignore_index=True)
xpl._case = "regression"
@@ -1207,7 +1007,7 @@ def test_contribution_plot_13(self):
xpl.postprocessing_modifications = False
xpl.preprocessing = None
# Creates a group of features named group1
- xpl.features_groups = {'group1': ['X1', 'X2']}
+ xpl.features_groups = {"group1": ["X1", "X2"]}
xpl.contributions_groups = xpl.state.compute_grouped_contributions(xpl.contributions, xpl.features_groups)
xpl.features_imp_groups = None
xpl._update_features_dict_with_groups(features_groups=xpl.features_groups)
@@ -1215,7 +1015,7 @@ def test_contribution_plot_13(self):
output = xpl.plot.contribution_plot(col, proba=False)
assert len(output.data) == 1
- assert output.data[0].type == 'scatter'
+ assert output.data[0].type == "scatter"
self.setUp()
def test_contribution_plot_14(self):
@@ -1223,15 +1023,11 @@ def test_contribution_plot_14(self):
contribution plot with groups of features for classification case and subset
"""
x_encoded = pd.DataFrame(
- data=np.array(
- [[0, 34],
- [1, 27]]),
- columns=['X1', 'X2'],
- index=['person_A', 'person_B']
+ data=np.array([[0, 34], [1, 27]]), columns=["X1", "X2"], index=["person_A", "person_B"]
)
xpl = self.smart_explainer
xpl.inv_features_dict = {}
- col = 'group1'
+ col = "group1"
xpl.x_encoded = x_encoded
xpl.contributions[0] = pd.concat([xpl.contributions[0]] * 10, ignore_index=True)
xpl.contributions[1] = pd.concat([xpl.contributions[1]] * 10, ignore_index=True)
@@ -1240,7 +1036,7 @@ def test_contribution_plot_14(self):
xpl.postprocessing_modifications = False
xpl.preprocessing = None
# Creates a group of features named group1
- xpl.features_groups = {'group1': ['X1', 'X2']}
+ xpl.features_groups = {"group1": ["X1", "X2"]}
xpl.contributions_groups = xpl.state.compute_grouped_contributions(xpl.contributions, xpl.features_groups)
xpl.features_imp_groups = None
xpl._update_features_dict_with_groups(features_groups=xpl.features_groups)
@@ -1249,7 +1045,7 @@ def test_contribution_plot_14(self):
output = xpl.plot.contribution_plot(col, proba=False, selection=subset)
assert len(output.data) == 1
- assert output.data[0].type == 'scatter'
+ assert output.data[0].type == "scatter"
assert len(output.data[0].x) == 10
self.setUp()
@@ -1258,15 +1054,11 @@ def test_contribution_plot_15(self):
contribution plot with groups of features for regression case with subset
"""
x_encoded = pd.DataFrame(
- data=np.array(
- [[0, 34],
- [1, 27]]),
- columns=['X1', 'X2'],
- index=['person_A', 'person_B']
+ data=np.array([[0, 34], [1, 27]]), columns=["X1", "X2"], index=["person_A", "person_B"]
)
xpl = self.smart_explainer
xpl.inv_features_dict = {}
- col = 'group1'
+ col = "group1"
xpl.x_encoded = x_encoded
xpl.contributions = pd.concat([self.contrib1] * 10, ignore_index=True)
xpl._case = "regression"
@@ -1277,7 +1069,7 @@ def test_contribution_plot_15(self):
xpl.postprocessing_modifications = False
xpl.preprocessing = None
# Creates a group of features named group1
- xpl.features_groups = {'group1': ['X1', 'X2']}
+ xpl.features_groups = {"group1": ["X1", "X2"]}
xpl.contributions_groups = xpl.state.compute_grouped_contributions(xpl.contributions, xpl.features_groups)
xpl.features_imp_groups = None
xpl._update_features_dict_with_groups(features_groups=xpl.features_groups)
@@ -1286,7 +1078,7 @@ def test_contribution_plot_15(self):
output = xpl.plot.contribution_plot(col, proba=False, selection=subset)
assert len(output.data) == 1
- assert output.data[0].type == 'scatter'
+ assert output.data[0].type == "scatter"
assert len(output.data[0].x) == 10
self.setUp()
@@ -1294,14 +1086,9 @@ def test_plot_features_import_1(self):
"""
Unit test plot features import 1
"""
- serie1 = pd.Series([0.131, 0.51], index=['col1', 'col2'])
+ serie1 = pd.Series([0.131, 0.51], index=["col1", "col2"])
output = self.smart_explainer.plot.plot_features_import(serie1)
- data = go.Bar(
- x=serie1,
- y=serie1.index,
- name='Global',
- orientation='h'
- )
+ data = go.Bar(x=serie1, y=serie1.index, name="Global", orientation="h")
expected_output = go.Figure(data=data)
assert np.array_equal(output.data[0].x, expected_output.data[0].x)
@@ -1313,21 +1100,11 @@ def test_plot_features_import_2(self):
"""
Unit test plot features import 2
"""
- serie1 = pd.Series([0.131, 0.51], index=['col1', 'col2'])
- serie2 = pd.Series([0.33, 0.11], index=['col1', 'col2'])
+ serie1 = pd.Series([0.131, 0.51], index=["col1", "col2"])
+ serie2 = pd.Series([0.33, 0.11], index=["col1", "col2"])
output = self.smart_explainer.plot.plot_features_import(serie1, serie2)
- data1 = go.Bar(
- x=serie1,
- y=serie1.index,
- name='Global',
- orientation='h'
- )
- data2 = go.Bar(
- x=serie2,
- y=serie2.index,
- name='Subset',
- orientation='h'
- )
+ data1 = go.Bar(x=serie1, y=serie1.index, name="Global", orientation="h")
+ data2 = go.Bar(x=serie2, y=serie2.index, name="Subset", orientation="h")
expected_output = go.Figure(data=[data2, data1])
assert np.array_equal(output.data[0].x, expected_output.data[0].x)
assert np.array_equal(output.data[0].y, expected_output.data[0].y)
@@ -1344,19 +1121,11 @@ def test_features_importance_1(self):
"""
xpl = self.smart_explainer
xpl.explain_data = None
- output = xpl.plot.features_importance(selection=['person_A', 'person_B'])
+ output = xpl.plot.features_importance(selection=["person_A", "person_B"])
- data1 = go.Bar(
- x=np.array([0.2296, 0.7704]),
- y=np.array(['Age', 'Education']),
- name='Subset',
- orientation='h')
+ data1 = go.Bar(x=np.array([0.2296, 0.7704]), y=np.array(["Age", "Education"]), name="Subset", orientation="h")
- data2 = go.Bar(
- x=np.array([0.2296, 0.7704]),
- y=np.array(['Age', 'Education']),
- name='Global',
- orientation='h')
+ data2 = go.Bar(x=np.array([0.2296, 0.7704]), y=np.array(["Age", "Education"]), name="Global", orientation="h")
expected_output = go.Figure(data=[data1, data2])
@@ -1374,25 +1143,17 @@ def test_features_importance_2(self):
Unit test features importance 2
"""
xpl = self.smart_explainer
- #regression
+ # regression
xpl.contributions = self.contrib1
xpl.backend.state = SmartState()
xpl.explain_data = None
xpl._case = "regression"
xpl.state = SmartState()
- output = xpl.plot.features_importance(selection=['person_A', 'person_B'])
+ output = xpl.plot.features_importance(selection=["person_A", "person_B"])
- data1 = go.Bar(
- x=np.array([0.2296, 0.7704]),
- y=np.array(['Age', 'Education']),
- name='Subset',
- orientation='h')
+ data1 = go.Bar(x=np.array([0.2296, 0.7704]), y=np.array(["Age", "Education"]), name="Subset", orientation="h")
- data2 = go.Bar(
- x=np.array([0.2296, 0.7704]),
- y=np.array(['Age', 'Education']),
- name='Global',
- orientation='h')
+ data2 = go.Bar(x=np.array([0.2296, 0.7704]), y=np.array(["Age", "Education"]), name="Global", orientation="h")
expected_output = go.Figure(data=[data1, data2])
@@ -1410,19 +1171,15 @@ def test_features_importance_3(self):
Unit test features importance for groups of features
"""
x_init = pd.DataFrame(
- data=np.array(
- [['PhD', 34, 1],
- ['Master', 27, 0]]),
- columns=['X1', 'X2', 'X3'],
- index=['person_A', 'person_B']
+ data=np.array([["PhD", 34, 1], ["Master", 27, 0]]),
+ columns=["X1", "X2", "X3"],
+ index=["person_A", "person_B"],
)
contrib = pd.DataFrame(
- data=np.array(
- [[-3.4, 0.78, 1.2],
- [1.2, 3.6, -0.3]]),
- columns=['X1', 'X2', 'X3'],
- index=['person_A', 'person_B']
+ data=np.array([[-3.4, 0.78, 1.2], [1.2, 3.6, -0.3]]),
+ columns=["X1", "X2", "X3"],
+ index=["person_A", "person_B"],
)
smart_explainer = SmartExplainer(model=self.model)
@@ -1431,10 +1188,10 @@ def test_features_importance_3(self):
smart_explainer.postprocessing_modifications = False
smart_explainer.features_imp_groups = None
smart_explainer.features_imp = None
- smart_explainer.features_groups = {'group0': ['X1', 'X2']}
+ smart_explainer.features_groups = {"group0": ["X1", "X2"]}
smart_explainer.contributions = [contrib, -contrib]
- smart_explainer.features_dict = {'X1': 'X1', 'X2': 'X2', 'X3': 'X3', 'group0': 'group0'}
- smart_explainer.inv_features_dict = {'X1': 'X1', 'X2': 'X2', 'X3': 'X3', 'group0': 'group0'}
+ smart_explainer.features_dict = {"X1": "X1", "X2": "X2", "X3": "X3", "group0": "group0"}
+ smart_explainer.inv_features_dict = {"X1": "X1", "X2": "X2", "X3": "X3", "group0": "group0"}
smart_explainer.model = self.smart_explainer.model
smart_explainer._case, smart_explainer._classes = check_model(self.smart_explainer.model)
smart_explainer.backend = ShapBackend(model=self.smart_explainer.model)
@@ -1442,17 +1199,15 @@ def test_features_importance_3(self):
smart_explainer.explain_data = None
smart_explainer.state = MultiDecorator(SmartState())
smart_explainer.contributions_groups = smart_explainer.state.compute_grouped_contributions(
- smart_explainer.contributions, smart_explainer.features_groups)
+ smart_explainer.contributions, smart_explainer.features_groups
+ )
smart_explainer.features_imp_groups = smart_explainer.state.compute_features_import(
- smart_explainer.contributions_groups)
+ smart_explainer.contributions_groups
+ )
output = smart_explainer.plot.features_importance()
- data1 = go.Bar(
- x=np.array([0.1682, 0.8318]),
- y=np.array(['X3', 'group0']),
- name='Global',
- orientation='h')
+ data1 = go.Bar(x=np.array([0.1682, 0.8318]), y=np.array(["X3", "group0"]), name="Global", orientation="h")
expected_output = go.Figure(data=[data1])
assert np.array_equal(output.data[0].x, expected_output.data[0].x)
@@ -1465,19 +1220,15 @@ def test_features_importance_4(self):
Unit test features importance for groups of features when displaying a group
"""
x_init = pd.DataFrame(
- data=np.array(
- [['PhD', 34, 1],
- ['Master', 27, 0]]),
- columns=['X1', 'X2', 'X3'],
- index=['person_A', 'person_B']
+ data=np.array([["PhD", 34, 1], ["Master", 27, 0]]),
+ columns=["X1", "X2", "X3"],
+ index=["person_A", "person_B"],
)
contrib = pd.DataFrame(
- data=np.array(
- [[-3.4, 0.78, 1.2],
- [1.2, 3.6, -0.3]]),
- columns=['X1', 'X2', 'X3'],
- index=['person_A', 'person_B']
+ data=np.array([[-3.4, 0.78, 1.2], [1.2, 3.6, -0.3]]),
+ columns=["X1", "X2", "X3"],
+ index=["person_A", "person_B"],
)
smart_explainer = SmartExplainer(model=self.model)
@@ -1486,10 +1237,10 @@ def test_features_importance_4(self):
smart_explainer.postprocessing_modifications = False
smart_explainer.features_imp_groups = None
smart_explainer.features_imp = None
- smart_explainer.features_groups = {'group0': ['X1', 'X2']}
+ smart_explainer.features_groups = {"group0": ["X1", "X2"]}
smart_explainer.contributions = [contrib, -contrib]
- smart_explainer.features_dict = {'X1': 'X1', 'X2': 'X2', 'X3': 'X3', 'group0': 'group0'}
- smart_explainer.inv_features_dict = {'X1': 'X1', 'X2': 'X2', 'X3': 'X3', 'group0': 'group0'}
+ smart_explainer.features_dict = {"X1": "X1", "X2": "X2", "X3": "X3", "group0": "group0"}
+ smart_explainer.inv_features_dict = {"X1": "X1", "X2": "X2", "X3": "X3", "group0": "group0"}
smart_explainer.model = self.smart_explainer.model
smart_explainer.backend = ShapBackend(model=self.smart_explainer.model)
smart_explainer.backend.state = MultiDecorator(SmartState())
@@ -1497,17 +1248,15 @@ def test_features_importance_4(self):
smart_explainer._case, smart_explainer._classes = check_model(self.smart_explainer.model)
smart_explainer.state = smart_explainer.backend.state
smart_explainer.contributions_groups = smart_explainer.state.compute_grouped_contributions(
- smart_explainer.contributions, smart_explainer.features_groups)
+ smart_explainer.contributions, smart_explainer.features_groups
+ )
smart_explainer.features_imp_groups = smart_explainer.state.compute_features_import(
- smart_explainer.contributions_groups)
+ smart_explainer.contributions_groups
+ )
- output = smart_explainer.plot.features_importance(group_name='group0')
+ output = smart_explainer.plot.features_importance(group_name="group0")
- data1 = go.Bar(
- x=np.array([0.4179, 0.4389]),
- y=np.array(['X2', 'X1']),
- name='Global',
- orientation='h')
+ data1 = go.Bar(x=np.array([0.4179, 0.4389]), y=np.array(["X2", "X1"]), name="Global", orientation="h")
expected_output = go.Figure(data=[data1])
assert np.array_equal(output.data[0].x, expected_output.data[0].x)
@@ -1517,7 +1266,7 @@ def test_features_importance_4(self):
def test_local_pred_1(self):
xpl = self.smart_explainer
- output = xpl.plot.local_pred('person_A',label=0)
+ output = xpl.plot.local_pred("person_A", label=0)
assert isinstance(output, float)
def test_plot_line_comparison_1(self):
@@ -1525,45 +1274,41 @@ def test_plot_line_comparison_1(self):
Unit test 1 for plot_line_comparison
"""
xpl = self.smart_explainer
- index = ['person_A', 'person_B']
- data = pd.DataFrame(
- data=np.array(
- [['PhD', 34],
- ['Master', 27]]
- ),
- columns=['X1', 'X2'],
- index=index
- )
- features_dict = {'X1': 'X1', 'X2': 'X2'}
- xpl.inv_features_dict = {'X1': 'X1', 'X2': 'X2'}
- colors = ['rgba(244, 192, 0, 1.0)', 'rgba(74, 99, 138, 0.7)']
- var_dict = ['X1', 'X2']
+ index = ["person_A", "person_B"]
+ data = pd.DataFrame(data=np.array([["PhD", 34], ["Master", 27]]), columns=["X1", "X2"], index=index)
+ features_dict = {"X1": "X1", "X2": "X2"}
+ xpl.inv_features_dict = {"X1": "X1", "X2": "X2"}
+ colors = ["rgba(244, 192, 0, 1.0)", "rgba(74, 99, 138, 0.7)"]
+ var_dict = ["X1", "X2"]
contributions = [[-3.4, 0.78], [1.2, 3.6]]
- title = 'Compare plot - index : person_A ; person_B '
+ title = "Compare plot - index : person_A ; person_B "
predictions = [data.loc[id] for id in index]
fig = list()
for i in range(2):
- fig.append(go.Scatter(
- x=[contributions[0][i], contributions[1][i]],
- y=['' + feat + ' ' for feat in var_dict],
- mode='lines+markers',
- name=f'Id: {index[i]} ',
- hovertext=[f'Id: {index[i]} X1 {index[i]} X2 " + feat + " " for feat in var_dict],
+ mode="lines+markers",
+ name=f"Id: {index[i]} ",
+ hovertext=[
+ f"Id: {index[i]} X1 {index[i]} X2 person_A ; person_B ' \
- + "person_A ; person_B "
+ + "person_B Education ', 'Age '],
- name='Id: person_B ',
- hovertext=['Id: person_B Education person_B Age Education ", "Age "],
+ name="Id: person_B ",
+ hovertext=[
+ "Id: person_B Education person_B Age person_A ;" \
- " person_B 0 ; person_B: 1 person_A ;"
+ " person_B 0 ; person_B: 1 Age ', 'Education '],
- name=f'Id: {index[i]} ',
- hovertext=[
- f'Id: {index[i]} Age {index[i]} Education Age ", "Education "],
+ name=f"Id: {index[i]} ",
+ hovertext=[
+ f"Id: {index[i]} Age {index[i]} Education A ; B A ; B Age ', 'Education '],
- name=f'Id: {index[i]} ',
- hovertext=[
- f'Id: {index[i]} Age {index[i]} Education Age ", "Education "],
+ name=f"Id: {index[i]} ",
+ hovertext=[
+ f"Id: {index[i]} Age {index[i]} Education Age ', 'Education '],
- name=f'Id: {index[i]} ',
- hovertext=[
- f'Id: {index[i]} Age {index[i]} Education Age ", "Education "],
+ name=f"Id: {index[i]} ",
+ hovertext=[
+ f"Id: {index[i]} Age {index[i]} Education 
- 
+ 
-As you can see, you can select on the right side which branch of your forked repository you want to associate to the pull request.
+As you can see, you can select on the right side which branch of your forked repository you want to associate to the pull request.
On the left side, you will find the official Shapash repository.
@@ -130,4 +130,4 @@ Your pull request is now ready to be submitted. A member of the Shapash team wil
You have contributed to an Open source project, thank you and congratulations ! 🥳
-Show your contribution to Shapash in your curriculum, and share it on your social media. Be proud of yourself, you gave some code lines to the entire world !
\ No newline at end of file
+Show your contribution to Shapash in your curriculum, and share it on your social media. Be proud of yourself, you gave some code lines to the entire world !
diff --git a/LICENSE b/LICENSE
index 4947287f..f433b1a5 100644
--- a/LICENSE
+++ b/LICENSE
@@ -174,4 +174,4 @@
incurred by, or claims asserted against, such Contributor by reason
of your accepting any such warranty or additional liability.
- END OF TERMS AND CONDITIONS
\ No newline at end of file
+ END OF TERMS AND CONDITIONS
diff --git a/Makefile b/Makefile
index 720b6909..1f65e00e 100644
--- a/Makefile
+++ b/Makefile
@@ -84,4 +84,4 @@ dist: clean ## builds source and wheel package
ls -l dist
install: clean ## install the package to the active Python's site-packages
- python setup.py install
\ No newline at end of file
+ python setup.py install
diff --git a/README.md b/README.md
index 8a159b3a..aee9e717 100644
--- a/README.md
+++ b/README.md
@@ -38,7 +38,7 @@ With Shapash, you can generate a **Webapp** that simplifies the comprehension of
Additionally, Shapash contributes to data science auditing by **presenting valuable information** about any model and data **in a comprehensive report**.
-Shapash is suitable for Regression, Binary Classification, and Multiclass problems. It is **compatible with numerous models**, including Catboost, Xgboost, LightGBM, Sklearn Ensemble, Linear models, and SVM. For other models, solutions to integrate Shapash are available; more details can be found [here](#how_shapash_works).
+Shapash is suitable for Regression, Binary Classification and Multiclass problems. It is **compatible with numerous models**, including Catboost, Xgboost, LightGBM, Sklearn Ensemble, Linear models, and SVM. For other models, solutions to integrate Shapash are available; more details can be found [here](#how_shapash_works).
> [!NOTE]
> If you want to give us feedback : [Feedback form](https://framaforms.org/shapash-collecting-your-feedback-and-use-cases-1687456776)
@@ -72,9 +72,9 @@ Shapash is suitable for Regression, Binary Classification, and Multiclass proble
| 2.2.x | Dataset Filter 
](https://github.com/MAIF/shapash/blob/master/tutorial/tutorial01-Shapash-Overview-Launch-WebApp.ipynb)
| 2.0.x | Refactoring Shapash 
](https://github.com/MAIF/shapash/blob/master/tutorial/explainer_and_backend/tuto-expl06-Shapash-custom-backend.ipynb)
| 1.7.x | Variabilize Colors 
](https://github.com/MAIF/shapash/blob/master/tutorial/common/tuto-common02-colors.ipynb)
-| 1.6.x | Explainability Quality Metrics 
](https://github.com/MAIF/shapash/blob/master/tutorial/explainability_quality/tuto-quality01-Builing-confidence-explainability.ipynb)
-| 1.4.x | Groups of features 
](https://github.com/MAIF/shapash/blob/master/tutorial/common/tuto-common01-groups_of_features.ipynb) |
-| 1.3.x | Shapash Report 
](https://github.com/MAIF/shapash/blob/master/tutorial/generate_report/tuto-shapash-report01.ipynb) |
+| 1.6.x | Explainability Quality Metrics 
- 
+ 
- 
+ 
' % s
@@ -136,4 +140,3 @@ def html_str_df_and_image(df: pd.DataFrame, fig: plt.Figure) -> str:
def print_figure(fig):
print_html(convert_fig_to_html(fig))
-
diff --git a/shapash/style/colors.json b/shapash/style/colors.json
index 513d9922..32999b6e 100644
--- a/shapash/style/colors.json
+++ b/shapash/style/colors.json
@@ -1,200 +1,200 @@
{
- "default": {
- "title_color": "rgb(50, 50, 50)",
- "axis_color": "rgb(50, 50, 50)",
- "featureimp_bar": {
- "1": "rgba(244, 192, 0, 1.0)",
- "2": "rgba(52, 55, 54, 0.7)"
- },
- "featureimp_line": "rgba(52, 55, 54, 0.8)",
- "featureimp_groups": {
- "0": "rgb(245, 133, 24)",
- "1": "rgb(49, 99, 149)"
- },
- "contrib_colorscale": [
- "rgb(52, 55, 54)",
- "rgb(74, 99, 138)",
- "rgb(116, 153, 214)",
- "rgb(162, 188, 213)",
- "rgb(212, 234, 242)",
- "rgb(235, 216, 134)",
- "rgb(255, 204, 83)",
- "rgb(244, 192, 0)",
- "rgb(255, 166, 17)",
- "rgb(255, 123, 38)",
- "rgb(255, 77, 7)"
- ],
- "violin_area_classif": {
- "0": "rgba(117, 152, 189, 0.9)",
- "1": "rgba(255, 166, 17, 0.9)"
- },
- "violin_default": "rgba(117, 152, 189, 0.9)",
- "localplot_bar": {
- "1": "rgba(244, 192, 0, 1.0)",
- "-1": "rgba(74, 99, 138, 0.7)",
- "0": "rgba(113, 101, 59, 1.0)",
- "-2": "rgba(52, 55, 54, 0.7)"
- },
- "localplot_line": {
- "1": "rgba(52, 55, 54, 0.8)",
- "-1": "rgba(27, 28, 28, 1.0)",
- "0": "rgba(52, 55, 54, 0.8)",
- "-2": "rgba(27, 28, 28, 1.0)"
- },
- "compare_plot": [
- "rgba(244, 192, 0, 1.0)",
- "rgba(74, 99, 138, 0.7)",
- "rgba(113, 101, 59, 1.0)",
- "rgba(183, 58, 56, 0.9)",
- "rgba(255, 123, 38, 1.0)",
- "rgba(0, 21, 179, 0.97)",
- "rgba(116, 1, 179, 0.9)"
- ],
- "interaction_scale": [
- "rgb(175, 169, 157)",
- "rgb(255, 255, 255)",
- "rgb(255, 77, 7)"
- ],
- "interaction_discrete": [
- "rgb(133, 92, 117)",
- "rgb(217, 175, 107)",
- "rgb(175, 100, 88)",
- "rgb(115, 111, 76)",
- "rgb(82, 106, 131)",
- "rgb(98, 83, 119)",
- "rgb(104, 133, 92)",
- "rgb(156, 156, 94)",
- "rgb(160, 97, 119)",
- "rgb(140, 120, 93)",
- "rgb(124, 124, 124)"
- ],
- "stability_bar": {
- "1": "rgba(255, 166, 17, 0.9)",
- "0": "rgba(117, 152, 189, 0.9)"
- },
- "compacity_bar": {
- "1": "rgba(255, 166, 17, 0.9)",
- "0": "rgba(117, 152, 189, 0.9)"
- },
- "webapp_button": {
- "0": "rgb(244, 192, 0)",
- "1": "rgb(113, 101, 59)"
- },
- "webapp_bkg": "rgb(52,55,54)",
- "webapp_title": "rgb(244, 192, 0)",
- "report_feature_distribution": {
- "train": "rgba(74, 99, 128, 0.7)",
- "test": "rgba(244, 192, 0, 1)",
- "true": "rgba(74, 99, 138, 0.7)",
- "pred": "rgba(244, 192, 0, 1)"
- },
- "report_confusion_matrix": [
- "rgb(255, 255, 255)",
- "rgb(244, 192, 0)",
- "rgb(255, 166, 17)"
- ],
- "prediction_plot": {
- "1": "rgba(52, 55, 54, 0.9)",
- "0": "rgba(255, 77, 7, 0.9)"
- }
- },
- "blues": {
- "title_color": "rgb(50, 50, 50)",
- "axis_color": "rgb(50, 50, 50)",
- "featureimp_bar": {
- "1": "rgba(0, 154, 203, 1)",
- "2": "rgba(223, 103, 0, 0.8)"
- },
- "featureimp_line": "rgba(52, 55, 54, 0.8)",
- "featureimp_groups": {
- "0": "rgb(10, 204, 143)",
- "1": "rgb(176, 140, 104)"
- },
- "contrib_colorscale": [
- "rgb(168, 84, 0)",
- "rgb(204, 102, 0)",
- "rgb(245, 122, 0)",
- "rgb(240, 150, 67)",
- "rgb(240, 195, 162)",
- "rgb(237, 235, 232)",
- "rgb(161, 221, 254)",
- "rgb(103, 208, 255)",
- "rgb(0, 154, 203)",
- "rgb(0, 98, 128)",
- "rgb(0, 70, 92)"
- ],
- "violin_area_classif": {
- "0": "rgba(223, 103, 0, 0.9)",
- "1": "rgba(0, 154, 203, 0.9)"
- },
- "violin_default": "rgba(0, 98, 128, 0.7)",
- "localplot_bar": {
- "1": "rgba(0, 154, 203, 1)",
- "-1": "rgba(235, 149, 72, 0.9)",
- "0": "rgba(106, 178, 204, 0.8)",
- "-2": "rgba(217, 116, 0, 0.5)"
- },
- "localplot_line": {
- "1": "rgba(52, 55, 54, 0.8)",
- "-1": "rgba(27, 28, 28, 1.0)",
- "0": "rgba(52, 55, 54, 0.8)",
- "-2": "rgba(27, 28, 28, 1.0)"
- },
- "compare_plot": [
- "rgba(0, 154, 203, 1.0)",
- "rgba(223, 103, 0, 0.8)",
- "rgba(11, 60, 217, 1.0)",
- "rgba(204, 50, 10, 0.9)",
- "rgba(11, 217, 152, 1.0)",
- "rgba(129, 64, 0, 0.9)",
- "rgba(0, 98, 128, 0.9)"
- ],
- "interaction_scale": [
- "rgb(175, 169, 157)",
- "rgb(255, 255, 255)",
- "rgb(0, 98, 128)"
- ],
- "interaction_discrete": [
- "rgb(95, 70, 144)",
- "rgb(29, 105, 150)",
- "rgb(56, 166, 165)",
- "rgb(15, 133, 84)",
- "rgb(115, 175, 72)",
- "rgb(237, 173, 8)",
- "rgb(225, 124, 5)",
- "rgb(204, 80, 62)",
- "rgb(148, 52, 110)",
- "rgb(111, 64, 112)",
- "rgb(102, 102, 102)"
- ],
- "stability_bar": {
- "1": "rgba(0, 154, 203, 0.9)",
- "0": "rgba(230, 111, 0, 0.9)"
- },
- "compacity_bar": {
- "1": "rgba(0, 154, 203, 0.9)",
- "0": "rgba(230, 111, 0, 0.9)"
- },
- "webapp_button": {
- "0": "rgb(0, 154, 203)",
- "1": "rgb(50, 90, 117)"
- },
- "webapp_bkg": "rgb(52,55,54)",
- "webapp_title": "rgb(0, 154, 203)",
- "report_feature_distribution": {
- "train": "rgba(0, 154, 203, 0.9)",
- "test": "rgba(230, 111, 0, 0.9)",
- "true": "rgba(0, 154, 203, 0.9)",
- "pred": "rgba(230, 111, 0, 0.9)"
- },
- "report_confusion_matrix": [
- "rgb(255, 255, 255)",
- "rgb(244, 192, 0)",
- "rgb(255, 166, 17)"
- ],
- "prediction_plot": {
- "1": "rgba(168, 84, 0, 0.9)",
- "0": "rgba(0, 70, 92, 0.9)"
- }
- }
-}
\ No newline at end of file
+ "blues": {
+ "axis_color": "rgb(50, 50, 50)",
+ "compacity_bar": {
+ "0": "rgba(230, 111, 0, 0.9)",
+ "1": "rgba(0, 154, 203, 0.9)"
+ },
+ "compare_plot": [
+ "rgba(0, 154, 203, 1.0)",
+ "rgba(223, 103, 0, 0.8)",
+ "rgba(11, 60, 217, 1.0)",
+ "rgba(204, 50, 10, 0.9)",
+ "rgba(11, 217, 152, 1.0)",
+ "rgba(129, 64, 0, 0.9)",
+ "rgba(0, 98, 128, 0.9)"
+ ],
+ "contrib_colorscale": [
+ "rgb(168, 84, 0)",
+ "rgb(204, 102, 0)",
+ "rgb(245, 122, 0)",
+ "rgb(240, 150, 67)",
+ "rgb(240, 195, 162)",
+ "rgb(237, 235, 232)",
+ "rgb(161, 221, 254)",
+ "rgb(103, 208, 255)",
+ "rgb(0, 154, 203)",
+ "rgb(0, 98, 128)",
+ "rgb(0, 70, 92)"
+ ],
+ "featureimp_bar": {
+ "1": "rgba(0, 154, 203, 1)",
+ "2": "rgba(223, 103, 0, 0.8)"
+ },
+ "featureimp_groups": {
+ "0": "rgb(10, 204, 143)",
+ "1": "rgb(176, 140, 104)"
+ },
+ "featureimp_line": "rgba(52, 55, 54, 0.8)",
+ "interaction_discrete": [
+ "rgb(95, 70, 144)",
+ "rgb(29, 105, 150)",
+ "rgb(56, 166, 165)",
+ "rgb(15, 133, 84)",
+ "rgb(115, 175, 72)",
+ "rgb(237, 173, 8)",
+ "rgb(225, 124, 5)",
+ "rgb(204, 80, 62)",
+ "rgb(148, 52, 110)",
+ "rgb(111, 64, 112)",
+ "rgb(102, 102, 102)"
+ ],
+ "interaction_scale": [
+ "rgb(175, 169, 157)",
+ "rgb(255, 255, 255)",
+ "rgb(0, 98, 128)"
+ ],
+ "localplot_bar": {
+ "-1": "rgba(235, 149, 72, 0.9)",
+ "-2": "rgba(217, 116, 0, 0.5)",
+ "0": "rgba(106, 178, 204, 0.8)",
+ "1": "rgba(0, 154, 203, 1)"
+ },
+ "localplot_line": {
+ "-1": "rgba(27, 28, 28, 1.0)",
+ "-2": "rgba(27, 28, 28, 1.0)",
+ "0": "rgba(52, 55, 54, 0.8)",
+ "1": "rgba(52, 55, 54, 0.8)"
+ },
+ "prediction_plot": {
+ "0": "rgba(0, 70, 92, 0.9)",
+ "1": "rgba(168, 84, 0, 0.9)"
+ },
+ "report_confusion_matrix": [
+ "rgb(255, 255, 255)",
+ "rgb(244, 192, 0)",
+ "rgb(255, 166, 17)"
+ ],
+ "report_feature_distribution": {
+ "pred": "rgba(230, 111, 0, 0.9)",
+ "test": "rgba(230, 111, 0, 0.9)",
+ "train": "rgba(0, 154, 203, 0.9)",
+ "true": "rgba(0, 154, 203, 0.9)"
+ },
+ "stability_bar": {
+ "0": "rgba(230, 111, 0, 0.9)",
+ "1": "rgba(0, 154, 203, 0.9)"
+ },
+ "title_color": "rgb(50, 50, 50)",
+ "violin_area_classif": {
+ "0": "rgba(223, 103, 0, 0.9)",
+ "1": "rgba(0, 154, 203, 0.9)"
+ },
+ "violin_default": "rgba(0, 98, 128, 0.7)",
+ "webapp_bkg": "rgb(52,55,54)",
+ "webapp_button": {
+ "0": "rgb(0, 154, 203)",
+ "1": "rgb(50, 90, 117)"
+ },
+ "webapp_title": "rgb(0, 154, 203)"
+ },
+ "default": {
+ "axis_color": "rgb(50, 50, 50)",
+ "compacity_bar": {
+ "0": "rgba(117, 152, 189, 0.9)",
+ "1": "rgba(255, 166, 17, 0.9)"
+ },
+ "compare_plot": [
+ "rgba(244, 192, 0, 1.0)",
+ "rgba(74, 99, 138, 0.7)",
+ "rgba(113, 101, 59, 1.0)",
+ "rgba(183, 58, 56, 0.9)",
+ "rgba(255, 123, 38, 1.0)",
+ "rgba(0, 21, 179, 0.97)",
+ "rgba(116, 1, 179, 0.9)"
+ ],
+ "contrib_colorscale": [
+ "rgb(52, 55, 54)",
+ "rgb(74, 99, 138)",
+ "rgb(116, 153, 214)",
+ "rgb(162, 188, 213)",
+ "rgb(212, 234, 242)",
+ "rgb(235, 216, 134)",
+ "rgb(255, 204, 83)",
+ "rgb(244, 192, 0)",
+ "rgb(255, 166, 17)",
+ "rgb(255, 123, 38)",
+ "rgb(255, 77, 7)"
+ ],
+ "featureimp_bar": {
+ "1": "rgba(244, 192, 0, 1.0)",
+ "2": "rgba(52, 55, 54, 0.7)"
+ },
+ "featureimp_groups": {
+ "0": "rgb(245, 133, 24)",
+ "1": "rgb(49, 99, 149)"
+ },
+ "featureimp_line": "rgba(52, 55, 54, 0.8)",
+ "interaction_discrete": [
+ "rgb(133, 92, 117)",
+ "rgb(217, 175, 107)",
+ "rgb(175, 100, 88)",
+ "rgb(115, 111, 76)",
+ "rgb(82, 106, 131)",
+ "rgb(98, 83, 119)",
+ "rgb(104, 133, 92)",
+ "rgb(156, 156, 94)",
+ "rgb(160, 97, 119)",
+ "rgb(140, 120, 93)",
+ "rgb(124, 124, 124)"
+ ],
+ "interaction_scale": [
+ "rgb(175, 169, 157)",
+ "rgb(255, 255, 255)",
+ "rgb(255, 77, 7)"
+ ],
+ "localplot_bar": {
+ "-1": "rgba(74, 99, 138, 0.7)",
+ "-2": "rgba(52, 55, 54, 0.7)",
+ "0": "rgba(113, 101, 59, 1.0)",
+ "1": "rgba(244, 192, 0, 1.0)"
+ },
+ "localplot_line": {
+ "-1": "rgba(27, 28, 28, 1.0)",
+ "-2": "rgba(27, 28, 28, 1.0)",
+ "0": "rgba(52, 55, 54, 0.8)",
+ "1": "rgba(52, 55, 54, 0.8)"
+ },
+ "prediction_plot": {
+ "0": "rgba(255, 77, 7, 0.9)",
+ "1": "rgba(52, 55, 54, 0.9)"
+ },
+ "report_confusion_matrix": [
+ "rgb(255, 255, 255)",
+ "rgb(244, 192, 0)",
+ "rgb(255, 166, 17)"
+ ],
+ "report_feature_distribution": {
+ "pred": "rgba(244, 192, 0, 1)",
+ "test": "rgba(244, 192, 0, 1)",
+ "train": "rgba(74, 99, 128, 0.7)",
+ "true": "rgba(74, 99, 138, 0.7)"
+ },
+ "stability_bar": {
+ "0": "rgba(117, 152, 189, 0.9)",
+ "1": "rgba(255, 166, 17, 0.9)"
+ },
+ "title_color": "rgb(50, 50, 50)",
+ "violin_area_classif": {
+ "0": "rgba(117, 152, 189, 0.9)",
+ "1": "rgba(255, 166, 17, 0.9)"
+ },
+ "violin_default": "rgba(117, 152, 189, 0.9)",
+ "webapp_bkg": "rgb(52,55,54)",
+ "webapp_button": {
+ "0": "rgb(244, 192, 0)",
+ "1": "rgb(113, 101, 59)"
+ },
+ "webapp_title": "rgb(244, 192, 0)"
+ }
+}
diff --git a/shapash/style/style_utils.py b/shapash/style/style_utils.py
index 4c722fec..36dad34c 100644
--- a/shapash/style/style_utils.py
+++ b/shapash/style/style_utils.py
@@ -9,23 +9,23 @@
def colors_loading():
"""
- colors_loading allows shapash to load a json file which contains different
+ colors_loading allows shapash to load a json file which contains different
palettes of colors that can be used in the plot
Returns
-------
- dict:
+ dict:
contains all available pallets
"""
current_path = os.path.dirname(os.path.abspath(__file__))
jsonfile = os.path.join(current_path, "colors.json")
- with open(jsonfile, 'r') as openfile:
+ with open(jsonfile) as openfile:
colors_dic = json.load(openfile)
return colors_dic
def select_palette(colors_dic, palette_name):
"""
- colors_loading allows shapash to load a json file which contains different
+ colors_loading allows shapash to load a json file which contains different
palettes of colors that can be used in the plot
Parameters
----------
@@ -35,7 +35,7 @@ def select_palette(colors_dic, palette_name):
name of the palette
Returns
-------
- dict:
+ dict:
contains colors of one palette
"""
if palette_name not in colors_dic.keys():
@@ -58,18 +58,18 @@ def convert_str_color_to_plt_format(txt):
>>> convert_str_color_to_plt_format(txt="rgba(244, 192, 0, 1)")
(0.96, 0.75, 0.0, 1.0)
"""
- txt = txt.replace('rgba', '').replace('rgb', '').replace('(', '').replace(')', '')
- list_txt = txt.split(',')
+ txt = txt.replace("rgba", "").replace("rgb", "").replace("(", "").replace(")", "")
+ list_txt = txt.split(",")
if len(list_txt) > 3:
- return [float(list_txt[i])/255 for i in range(3)] + [float(list_txt[3])]
+ return [float(list_txt[i]) / 255 for i in range(3)] + [float(list_txt[3])]
else:
- return [float(x)/255 for x in list_txt]
+ return [float(x) / 255 for x in list_txt]
def define_style(palette):
"""
- the define_style function is a function that uses a palette
- to define the different styles used in the different outputs
+ the define_style function is a function that uses a palette
+ to define the different styles used in the different outputs
of Shapash
Parameters
----------
@@ -77,111 +77,53 @@ def define_style(palette):
contains colors of one palette
Returns
-------
- dict :
+ dict :
contains different style elements
"""
style_dict = dict()
- style_dict['dict_title'] = {
- 'xanchor': "center",
- 'yanchor': "middle",
- 'x': 0.5,
- 'y': 0.9,
- 'font': {
- 'size': 24,
- 'family': "Arial",
- 'color': palette["title_color"]
- }
+ style_dict["dict_title"] = {
+ "xanchor": "center",
+ "yanchor": "middle",
+ "x": 0.5,
+ "y": 0.9,
+ "font": {"size": 24, "family": "Arial", "color": palette["title_color"]},
}
- style_dict['dict_title_stability'] = {
- 'xanchor': "center",
- 'x': 0.5,
+ style_dict["dict_title_stability"] = {
+ "xanchor": "center",
+ "x": 0.5,
"yanchor": "bottom",
"pad": dict(b=50),
- 'font': {
- 'size': 24,
- 'family': "Arial",
- 'color': palette["title_color"]
- }
+ "font": {"size": 24, "family": "Arial", "color": palette["title_color"]},
}
featureimp_bar = convert_string_to_int_keys(palette["featureimp_bar"])
- style_dict['dict_featimp_colors'] = {
- 1: {
- 'color': featureimp_bar[1],
- 'line': {
- 'color': palette["featureimp_line"],
- 'width': 0.5
- }
- },
- 2: {
- 'color': featureimp_bar[2]
- }
- }
- style_dict['featureimp_groups'] = list(palette["featureimp_groups"].values())
- style_dict['init_contrib_colorscale'] = palette["contrib_colorscale"]
- style_dict['violin_area_classif'] = list(palette["violin_area_classif"].values())
- style_dict['prediction_plot'] = list(palette["prediction_plot"].values())
- style_dict['violin_default'] = palette["violin_default"]
- style_dict['dict_title_compacity'] = {
- 'font': {
- 'size': 14,
- 'family': "Arial",
- 'color': palette["title_color"]
- }
- }
- style_dict['dict_xaxis'] = {
- 'font': {
- 'size': 16,
- 'family': "Arial Black",
- 'color': palette["axis_color"]
- }
- }
- style_dict['dict_yaxis'] = {
- 'font': {
- 'size': 16,
- 'family': "Arial Black",
- 'color': palette["axis_color"]
- }
+ style_dict["dict_featimp_colors"] = {
+ 1: {"color": featureimp_bar[1], "line": {"color": palette["featureimp_line"], "width": 0.5}},
+ 2: {"color": featureimp_bar[2]},
}
+ style_dict["featureimp_groups"] = list(palette["featureimp_groups"].values())
+ style_dict["init_contrib_colorscale"] = palette["contrib_colorscale"]
+ style_dict["violin_area_classif"] = list(palette["violin_area_classif"].values())
+ style_dict["prediction_plot"] = list(palette["prediction_plot"].values())
+ style_dict["violin_default"] = palette["violin_default"]
+ style_dict["dict_title_compacity"] = {"font": {"size": 14, "family": "Arial", "color": palette["title_color"]}}
+ style_dict["dict_xaxis"] = {"font": {"size": 16, "family": "Arial Black", "color": palette["axis_color"]}}
+ style_dict["dict_yaxis"] = {"font": {"size": 16, "family": "Arial Black", "color": palette["axis_color"]}}
localplot_bar = convert_string_to_int_keys(palette["localplot_bar"])
localplot_line = convert_string_to_int_keys(palette["localplot_line"])
- style_dict['dict_local_plot_colors'] = {
- 1: {
- 'color': localplot_bar[1],
- 'line': {
- 'color': localplot_line[1],
- 'width': 0.5
- }
- },
- -1: {
- 'color':localplot_bar[-1],
- 'line': {
- 'color': localplot_line[-1],
- 'width': 0.5
- }
- },
- 0: {
- 'color': localplot_bar[0],
- 'line': {
- 'color': localplot_line[0],
- 'width': 0.5
- }
- },
- -2: {
- 'color': localplot_bar[-2],
- 'line': {
- 'color': localplot_line[-2],
- 'width': 0.5
- }
- }
+ style_dict["dict_local_plot_colors"] = {
+ 1: {"color": localplot_bar[1], "line": {"color": localplot_line[1], "width": 0.5}},
+ -1: {"color": localplot_bar[-1], "line": {"color": localplot_line[-1], "width": 0.5}},
+ 0: {"color": localplot_bar[0], "line": {"color": localplot_line[0], "width": 0.5}},
+ -2: {"color": localplot_bar[-2], "line": {"color": localplot_line[-2], "width": 0.5}},
}
- style_dict['dict_compare_colors'] = palette["compare_plot"]
- style_dict['interactions_col_scale'] = palette["interaction_scale"]
- style_dict['interactions_discrete_colors'] = palette["interaction_discrete"]
- style_dict['dict_stability_bar_colors'] = convert_string_to_int_keys(palette["stability_bar"])
- style_dict['dict_compacity_bar_colors'] = convert_string_to_int_keys(palette["compacity_bar"])
- style_dict['webapp_button'] = convert_string_to_int_keys(palette["webapp_button"])
- style_dict['webapp_bkg'] = palette["webapp_bkg"]
- style_dict['webapp_title'] = palette["webapp_title"]
+ style_dict["dict_compare_colors"] = palette["compare_plot"]
+ style_dict["interactions_col_scale"] = palette["interaction_scale"]
+ style_dict["interactions_discrete_colors"] = palette["interaction_discrete"]
+ style_dict["dict_stability_bar_colors"] = convert_string_to_int_keys(palette["stability_bar"])
+ style_dict["dict_compacity_bar_colors"] = convert_string_to_int_keys(palette["compacity_bar"])
+ style_dict["webapp_button"] = convert_string_to_int_keys(palette["webapp_button"])
+ style_dict["webapp_bkg"] = palette["webapp_bkg"]
+ style_dict["webapp_title"] = palette["webapp_title"]
return style_dict
@@ -218,10 +160,7 @@ def get_pyplot_color(colors):
if isinstance(colors, str):
return convert_str_color_to_plt_format(colors)
elif isinstance(colors, dict):
- dict_color_palette = {
- k: convert_str_color_to_plt_format(v)
- for k, v in colors.items()
- }
+ dict_color_palette = {k: convert_str_color_to_plt_format(v) for k, v in colors.items()}
return dict_color_palette
elif isinstance(colors, list):
return [convert_str_color_to_plt_format(v) for v in colors]
diff --git a/shapash/utils/category_encoder_backend.py b/shapash/utils/category_encoder_backend.py
index aea24aae..1ce95c15 100644
--- a/shapash/utils/category_encoder_backend.py
+++ b/shapash/utils/category_encoder_backend.py
@@ -2,9 +2,9 @@
Category_encoder
"""
-import pandas as pd
-import numpy as np
import category_encoders as ce
+import numpy as np
+import pandas as pd
category_encoder_onehot = "