Fix tests, README links

.github/workflows/test.yml

@@ -53,9 +53,15 @@ jobs:
           python-version: ${{ matrix.python-version }}
           cache: "pip" # Caching pip dependencies.
       - name: Install dependencies
+        if: matrix.python-version != '3.7'
         run: |
           pip install --upgrade pip
           pip install .[dev]
+      - name: Install dependencies (python 3.7, use legacy resolver)
+        if: matrix.python-version == '3.7'
+        run: |
+          pip install --upgrade pip
+          pip install .[dev] --use-deprecated=legacy-resolver
       - name: Test with pytest (xdist)
         if: matrix.os != 'ubuntu-latest'
         run: pytest -vvvx -n auto -m "not slow and not extra and not skipci" --durations=50 --cov

.github/workflows/test_full.yml

@@ -49,9 +49,15 @@ jobs:
           python-version: ${{ matrix.python-version }}
           cache: "pip" # Caching pip dependencies.
       - name: Install dependencies
+        if: matrix.python-version != '3.7'
         run: |
           pip install --upgrade pip
           pip install .[dev]
+      - name: Install dependencies (python 3.7, use legacy resolver)
+        if: matrix.python-version == '3.7'
+        run: |
+          pip install --upgrade pip
+          pip install .[dev] --use-deprecated=legacy-resolver
       - name: Test with pytest (xdist)
         if: matrix.os != 'ubuntu-latest'
         run: pytest -vvvx -n auto -m "not skipci" --durations=50 --cov

.pre-commit-config.yaml

@@ -1,4 +1,4 @@
-exclude: '^docs/conf.py'
+# exclude: '^docs/conf.py' # If need global exclude.
 
 repos:
 - repo: https://github.com/pre-commit/pre-commit-hooks
@@ -22,6 +22,7 @@ repos:
     exclude: 'docs/user_guide'  # Auto-generated, leave as is.
   - id: pretty-format-json
     args: ['--autofix', '--indent=4', '--no-sort-keys']
+    exclude: '.*.ipynb'  # ipynb may get treated as JSON, don't want that.
   - id: requirements-txt-fixer
     exclude: docs/requirements.txt
   - id: mixed-line-ending
@@ -33,39 +34,34 @@ repos:
   - id: isort
 
 - repo: https://github.com/psf/black
-  rev: 22.12.0
+  rev: 23.7.0
   hooks:
   - id: black-jupyter
     language_version: python3
-    args: [
-      "--force-exclude=docs/conf.py"
-    ]
 
 # If like to embrace black styles even in the docs:
 - repo: https://github.com/asottile/blacken-docs
-  rev: v1.12.1
+  rev: 1.16.0
   hooks:
   - id: blacken-docs
     additional_dependencies: [black]
     exclude: 'docs/user_guide'  # Auto-generated, leave as is.
 
 - repo: https://github.com/PyCQA/flake8
-  rev: 6.0.0
+  rev: 6.1.0
   hooks:
   - id: flake8
-  ## You can add flake8 plugins via `additional_dependencies`:
-  #  additional_dependencies: [flake8-bugbear]
 
 -   repo: https://github.com/pre-commit/mirrors-mypy
-    rev: v1.1.1
+    rev: v1.5.1
     hooks:
     -   id: mypy
         files: "^src/"
         args: [--config-file=mypy.ini]
-        additional_dependencies: [pydantic]
+        additional_dependencies: [pydantic<2]
 
 - repo: https://github.com/PyCQA/bandit
-  rev: 1.7.4
+  rev: 1.7.5
   hooks:
   - id: bandit
     args: ["-c", "pyproject.toml", "-q"]

README.md

@@ -6,7 +6,7 @@
 -->
 
 <!-- exclude_docs -->
-[![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/user_guide/tutorial04_prediction.ipynb)
+[![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/usage/tutorial04_prediction.ipynb)
 [![Documentation Status](https://readthedocs.org/projects/temporai/badge/?version=latest)](https://temporai.readthedocs.io/en/latest/?badge=latest)
 
 [![Python 3.7+](https://img.shields.io/badge/python-3.7+-blue.svg)](https://www.python.org/downloads/release/python-370/)
@@ -423,15 +423,15 @@ Prediction where targets are temporal (time series).
 - [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/data/tutorial04_data_splitting.ipynb) - [Data Splitting](./tutorials/data/tutorial04_data_splitting.ipynb)
 
 ### User Guide
-- [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/user_guide/tutorial01_plugins.ipynb) - [Plugins](./tutorials/user_guide/tutorial01_plugins.ipynb)
-- [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/user_guide/tutorial02_imputation.ipynb) - [Imputation](./tutorials/user_guide/tutorial02_imputation.ipynb)
-- [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/user_guide/tutorial03_scaling.ipynb) - [Scaling](./tutorials/user_guide/tutorial03_scaling.ipynb)
-- [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/user_guide/tutorial04_prediction.ipynb) - [Prediction](./tutorials/user_guide/tutorial04_prediction.ipynb)
-- [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/user_guide/tutorial05_time_to_event.ipynb) - [Time-to-event Analysis](./tutorials/user_guide/tutorial05_time_to_event.ipynb)
-- [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/user_guide/tutorial06_treatments.ipynb) - [Treatment Effects](./tutorials/user_guide/tutorial06_treatments.ipynb)
-- [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/user_guide/tutorial07_pipeline.ipynb) - [Pipeline](./tutorials/user_guide/tutorial07_pipeline.ipynb)
-- [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/user_guide/tutorial08_benchmarks.ipynb) - [Benchmarks](./tutorials/user_guide/tutorial08_benchmarks.ipynb)
-- [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/user_guide/tutorial09_automl.ipynb) - [AutoML](./tutorials/user_guide/tutorial09_automl.ipynb)
+- [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/usage/tutorial01_plugins.ipynb) - [Plugins](./tutorials/usage/tutorial01_plugins.ipynb)
+- [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/usage/tutorial02_imputation.ipynb) - [Imputation](./tutorials/usage/tutorial02_imputation.ipynb)
+- [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/usage/tutorial03_scaling.ipynb) - [Scaling](./tutorials/usage/tutorial03_scaling.ipynb)
+- [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/usage/tutorial04_prediction.ipynb) - [Prediction](./tutorials/usage/tutorial04_prediction.ipynb)
+- [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/usage/tutorial05_time_to_event.ipynb) - [Time-to-event Analysis](./tutorials/usage/tutorial05_time_to_event.ipynb)
+- [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/usage/tutorial06_treatments.ipynb) - [Treatment Effects](./tutorials/usage/tutorial06_treatments.ipynb)
+- [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/usage/tutorial07_pipeline.ipynb) - [Pipeline](./tutorials/usage/tutorial07_pipeline.ipynb)
+- [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/usage/tutorial08_benchmarks.ipynb) - [Benchmarks](./tutorials/usage/tutorial08_benchmarks.ipynb)
+- [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/usage/tutorial09_automl.ipynb) - [AutoML](./tutorials/usage/tutorial09_automl.ipynb)
 
 ### Extending TemporAI
 - [![Test In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/vanderschaarlab/temporai/blob/main/tutorials/extending/tutorial01_custom_plugin.ipynb) - [Writing a Custom Plugin](./tutorials/extending/tutorial01_custom_plugin.ipynb)

pytest.ini

@@ -24,6 +24,14 @@ filterwarnings =
     ignore:.*rsyncdir.*:DeprecationWarning
     # Jupyter platformdirs-related warning, ignore:
     ignore:.*platformdirs.*:DeprecationWarning
+    # Some pandas FutureWarnings that can be ignored for now:
+    ignore:.*is_sparse.*:FutureWarning
+    ignore:.*is_categorical_dtype.*:FutureWarning
+    ignore:.*swapaxes.*:FutureWarning
+    # Some numpy DeprecationWarnings:
+    ignore:.*bool8.*:DeprecationWarning
+    # DeprecationWarning about pkg_resources (triggered by lifelines):
+    ignore:.*pkg_resources.*:DeprecationWarning
 # Use pytest markers to select/deselect specific tests
 markers =
     internet: requires internet access

setup.cfg

@@ -56,8 +56,6 @@ python_requires = >=3.7
 # new major versions. This works if the required packages follow Semantic Versioning.
 # For more information, check out https://semver.org/.
 install_requires =
-    # catboost 1.2+ has build errors on "macos-latest" GH runners: https://github.com/catboost/catboost/issues/2371
-    catboost < 1.2
     clairvoyance2 >=0.0.2
     cloudpickle
     geomloss>=0.2.6
@@ -82,8 +80,10 @@ install_requires =
     torchdiffeq
     torchlaplace >= 0.0.4
     tsai
-    typing-extensions
+    typing-extensions >= 4.7.1
     xgbse
+    # joblib has a bug with py37 & Windows: https://github.com/joblib/loky/issues/411
+    joblib < 1.3.0; python_version=="3.7" and platform_system=="Windows"
 
 
 [options.packages.find]

src/tempor/benchmarks/benchmark.py

@@ -132,7 +132,6 @@ def benchmark_models(
 
 @pydantic.validate_arguments(config=dict(arbitrary_types_allowed=True))
 def visualize_benchmark(results: Dict[str, pd.DataFrame], palette: str = "viridis") -> Any:
-
     # Pre-format DF for plotting.
     for k, v in results.items():
         v["method"] = k

src/tempor/benchmarks/evaluation.py

@@ -332,7 +332,6 @@ def roc_auc_score(self, y_test: np.ndarray, y_pred_proba: np.ndarray) -> float:
         return utils.evaluate_auc_multiclass(y_test, y_pred_proba)[0]
 
     def average_precision_score(self, y_test: np.ndarray, y_pred_proba: np.ndarray) -> float:
-
         return utils.evaluate_auc_multiclass(y_test, y_pred_proba)[1]
 
 

src/tempor/benchmarks/utils.py

@@ -38,7 +38,6 @@ def evaluate_auc_multiclass(
     y_pred_proba_tmp = get_y_pred_proba_hlpr(y_pred_proba, n_classes)
 
     if n_classes > 2:
-
         log.debug(f"+evaluate_auc {y_test.shape} {y_pred_proba_tmp.shape}")
 
         fpr = dict()
@@ -60,7 +59,6 @@ def evaluate_auc_multiclass(
         aucroc = roc_auc["micro"]
         aucprc = average_precision["micro"]
     else:
-
         aucroc = roc_auc_score(np.ravel(y_test), y_pred_proba_tmp, multi_class="ovr")
         aucprc = average_precision_score(np.ravel(y_test), y_pred_proba_tmp)
 

src/tempor/data/samples.py

@@ -3,12 +3,14 @@
 # pylint: disable=unnecessary-ellipsis
 
 import abc
+import contextlib
 from typing import TYPE_CHECKING, Any, List, Optional, Tuple
 
 import numpy as np
 import pandas as pd
 import pandera as pa
 import pydantic
+from packaging.version import Version
 from typing_extensions import Self
 
 import tempor.exc
@@ -292,6 +294,35 @@ def __getitem__(self, key: data_typing.GetItemKey) -> Self:
         )
 
 
+@contextlib.contextmanager
+def workaround_pandera_pd2_1_0_multiindex_compatibility(schema: pa.DataFrameSchema, data: pd.DataFrame):
+    """A version compatibility issue exists between pandera and pandas 2.1.0, as reported here:
+    https://github.com/unionai-oss/pandera/issues/1328
+
+    The error pertains to multiindex uniqueness validation giving an unexpected error.
+
+    This is a workaround that will "manually" throw an error that is expected from pandera.
+    """
+
+    def problem_versions() -> bool:
+        return Version(pd.__version__) >= Version("2.1.0")
+        # TODO: When/if fixed in pandera, add the below condition:
+        # and Version(pa.__version__) < Version("0.XX.YY")
+
+    try:
+        yield
+
+    except ValueError as ex:
+        if problem_versions() and "Columns with duplicate values are not supported in stack" in str(ex):
+            cols = data.index.names
+            raise pa.errors.SchemaError(schema=schema, data=data, message=f"columns {cols} not unique")
+        else:
+            raise
+
+    finally:
+        pass
+
+
 class TimeSeriesSamples(DataSamples):
     _data: pd.DataFrame
     _schema: pa.DataFrameSchema
@@ -397,7 +428,8 @@ def _validate(self) -> None:
             coerce=False,
             unique=multiindex_unique_def,
         )
-        self._data = schema.validate(data)
+        with workaround_pandera_pd2_1_0_multiindex_compatibility(schema, data):
+            self._data = schema.validate(data)
 
         logger.debug(f"Final schema:\n{schema}")
         self._schema = schema

src/tempor/models/ddh.py

@@ -80,7 +80,6 @@ def __init__(
         clipping_value: int = 1,
         output_mode: str = "MLP",
     ) -> None:
-
         self.split = split
         self.split_time = None
 

src/tempor/utils/dataloaders/sine.py

@@ -89,15 +89,13 @@ def load(self, **kwargs) -> dataset.OneOffPredictionDataset:
         # Generate sine data.
 
         for i in range(self.no):
-
             # Initialize each time-series
             local = list()
 
             # For each feature
             seq_len = self.seq_len
 
             for k in range(self.temporal_dim):  # pylint: disable=unused-variable
-
                 # Randomly drawn frequency and phase:
                 freq = np.random.beta(2, 2)
                 phase = np.random.normal()

tests/automl/test_pipeline_selector.py

@@ -185,7 +185,6 @@ def assert_actual_params(pipe, sample, method):
         assert getattr(estimator.params, k) == v
 
 
-@pytest.mark.filterwarnings("ignore:.*bool8.*:DeprecationWarning")  # Expected.
 @pytest.mark.parametrize("task_type,predictor", TEST_PREDICTOR_CASES)
 @pytest.mark.parametrize("static_imputers", TEST_STATIC_IMPUTERS_CASES)
 @pytest.mark.parametrize("static_scalers", TEST_STATIC_SCALERS_CASES)

tests/automl/test_seeker.py

@@ -73,7 +73,6 @@ def patched_transform(self, *args, **kwargs):  # pylint: disable=unused-argument
     np.random.seed(12345)
 
     def patched_evaluate(*args, **kwargs):
-
         # Since "ensure reproducibility" may affect seeding, and we want this function to return different values,
         # seed it from input hash manually as below.
         seed_from = str(args[0])
@@ -610,7 +609,6 @@ def test_init_fails_wrong_overrides(self, get_dataset: Callable):
                 override_hp_space=override_hp_space,  # type: ignore
             )
 
-    @pytest.mark.filterwarnings("ignore:.*bool8.*:DeprecationWarning")  # Expected.
     @pytest.mark.parametrize(
         "task_type,estimator_names,metric",
         [

tests/conftest.py

@@ -191,6 +191,7 @@ def func(ds: "dataset.PredictiveDataset") -> "dataset.CovariatesDataset":
 
 # --- Reusable datasets. ---
 
+
 # Sine data: full.
 @pytest.fixture(scope="session")
 def _sine_data_full():

tests/data/test_samples.py

@@ -29,7 +29,6 @@ class DFsTest:
 
 
 def set_up_dfs_test():
-
     # --- Static. ---
 
     categories = ["A", "B", "C"]

tests/log/test_custom_logger.py

@@ -73,7 +73,6 @@ def as_loguru_logs(records: List[Any]) -> str:
 
 
 def test_console_logging_at_trace(caplog):  # noqa: F811
-
     config = tempor.get_config()
     config.logging.level = "TRACE"
     tempor.configure(config)
@@ -97,7 +96,6 @@ def test_console_logging_at_trace(caplog):  # noqa: F811
 
 
 def test_console_logging_at_info(caplog):  # noqa: F811
-
     config = tempor.get_config()
     config.logging.level = "INFO"
     tempor.configure(config)
@@ -121,7 +119,6 @@ def test_console_logging_at_info(caplog):  # noqa: F811
 
 
 def test_console_logging_at_error(caplog):  # noqa: F811
-
     config = tempor.get_config()
     config.logging.level = "ERROR"
     tempor.configure(config)

tests/plugins/preprocessing/imputation/static/test_static_tabular_imputer.py

@@ -91,8 +91,6 @@ def test_transform(
     assert output.static.dataframe().isna().sum().sum() == 0
 
 
-@pytest.mark.filterwarnings("ignore:.*Jupyter.*platformdirs.*:DeprecationWarning")  # Expected.
-@pytest.mark.filterwarnings("ignore:.*bool8.*:DeprecationWarning")  # Expected.
 @pytest.mark.filterwarnings("ignore:.*nonzero.*0d.*:DeprecationWarning")  # Expected for EM imputer.
 @pytest.mark.filterwarnings("ignore::RuntimeWarning")  # Expected for EM imputer.
 @pytest.mark.filterwarnings("ignore::sklearn.exceptions.ConvergenceWarning")  # May happen in some cases.

tests/plugins/preprocessing/imputation/temporal/test_ts_tabular_imputer.py

@@ -84,8 +84,6 @@ def test_transform(
     assert output.time_series.dataframe().isna().sum().sum() == 0
 
 
-@pytest.mark.filterwarnings("ignore:.*Jupyter.*platformdirs.*:DeprecationWarning")  # Expected.
-@pytest.mark.filterwarnings("ignore:.*bool8.*:DeprecationWarning")  # Expected.
 @pytest.mark.filterwarnings("ignore:.*nonzero.*0d.*:DeprecationWarning")  # Expected for EM imputer.
 @pytest.mark.filterwarnings("ignore::RuntimeWarning")  # Expected for EM imputer.
 @pytest.mark.filterwarnings("ignore::sklearn.exceptions.ConvergenceWarning")  # May happen in some cases.

tests/plugins/test_all_plugins.py

@@ -16,7 +16,6 @@ def test_init_success(plugin_fqn):
 
 
 # DeprecationWarning expected for preprocessing.imputation.temporal.ts_tabular_imputer:
-@pytest.mark.filterwarnings("ignore:.*bool8.*:DeprecationWarning")
 @pytest.mark.parametrize("plugin_fqn", PLUGIN_FQNS)
 def test_sample_hyperparameters(plugin_fqn):
     PluginCls = plugin_loader.get_class(plugin_fqn)