Merge branch 'master' of https://github.com/casact/chainladder-python …

…into 497

chainladder/core/base.py

@@ -2,6 +2,8 @@
 # License, v. 2.0. If a copy of the MPL was not distributed with this
 # file, You can obtain one at https://mozilla.org/MPL/2.0/.
 import pandas as pd
+from packaging import version
+
 import numpy as np
 from chainladder.utils.cupy import cp
 from chainladder.utils.sparse import sp
@@ -402,7 +404,30 @@ def __array_ufunc__(self, ufunc, method, *inputs, **kwargs):
             return obj
         else:
             raise NotImplementedError()
-
+
+    def _interchange_dataframe(self, data):
+        """
+        Convert an object supporting the __dataframe__ protocol to a pandas DataFrame.
+        Requires pandas version > 1.5.2.
+
+        Parameters
+        ----------
+        data :
+            Dataframe object supporting the __dataframe__ protocol.
+        Returns
+        -------
+        out: pd.DataFrame
+            A pandas DataFrame.
+        """
+        # Check if pandas version is greater than 1.5.2
+        if version.parse(pd.__version__) >= version.parse("1.5.2"):
+            return pd.api.interchange.from_dataframe(data)
+
+        else:
+            # Raise an error prompting the user to upgrade pandas
+            raise NotImplementedError("Your version of pandas does not support the DataFrame interchange API. "
+                                    "Please upgrade pandas to a version greater than 1.5.2 to use this feature.")
+
     def __array_function__(self, func, types, args, kwargs):
         from chainladder.utils.utility_functions import concat
 

chainladder/core/tests/test_triangle.py

@@ -1,5 +1,6 @@
 import chainladder as cl
 import pandas as pd
+import polars as pl
 import numpy as np
 import copy
 import pytest
@@ -745,6 +746,9 @@ def test_halfyear_development():
         ["2012-01-01", "2013-12-31", "incurred", 200.0],
     ]
 
+    df_polars = pl.DataFrame(data)
+    df_polars.columns = ["origin", "val_date", "idx", "value"]
+
     assert (
         type(
             cl.Triangle(
@@ -757,3 +761,35 @@ def test_halfyear_development():
             )
         )
     ) == cl.Triangle
+
+    assert (
+        type(
+            cl.Triangle(
+                data=df_polars,
+                index="idx",
+                columns="value",
+                origin="origin",
+                development="val_date",
+                cumulative=True,
+            )
+        )
+    ) == cl.Triangle
+
+    assert (
+            cl.Triangle(
+                data=pd.DataFrame(data, columns=["origin", "val_date", "idx", "value"]),
+                index="idx",
+                columns="value",
+                origin="origin",
+                development="val_date",
+                cumulative=True,
+            ) ==
+            cl.Triangle(
+                data=df_polars,
+                index="idx",
+                columns="value",
+                origin="origin",
+                development="val_date",
+                cumulative=True,
+            )
+    )

chainladder/core/triangle.py

@@ -123,7 +123,9 @@ def __init__(
     ):
         if data is None:
             return
-
+        elif not isinstance(data, pd.DataFrame) and hasattr(data, "__dataframe__"):
+            data = self._interchange_dataframe(data)
+
         index, columns, origin, development = self._input_validation(
             data, index, columns, origin, development
         )
@@ -274,7 +276,7 @@ def __init__(
             self.ddims = obj.ddims
             self.values = obj.values
             self.valuation_date = pd.Timestamp(options.ULT_VAL)
-
+    
     @staticmethod
     def _split_ult(data, index, columns, origin, development):
         """Deal with triangles with ultimate values"""
@@ -804,7 +806,7 @@ def broadcast_axis(self, axis, value):
         return self
 
     def copy(self):
-        X = Triangle()
+        X = object.__new__(self.__class__)
         X.__dict__.update(vars(self))
         X._set_slicers()
         X.values = X.values.copy()

chainladder/methods/chainladder.py

@@ -49,7 +49,7 @@ def fit(self, X, y=None, sample_weight=None):
         return self
 
     def predict(self, X, sample_weight=None):
-        """Predicts the Benktander ultimate on a new triangle **X**
+        """Predicts the chainladder ultimate on a new triangle **X**
 
         Parameters
         ----------
@@ -61,7 +61,7 @@ def predict(self, X, sample_weight=None):
         Returns
         -------
         X_new: Triangle
-            Loss data with Benktander ultimate applied
+            Loss data with chainladder ultimate applied
         """
         X_new = super().predict(X, sample_weight)
         X_new.ultimate_ = self._get_ultimate(X_new, sample_weight)

chainladder/methods/expectedloss.py

@@ -39,7 +39,7 @@ def __init__(self, apriori=1.0, apriori_sigma=0.0, random_state=None):
         self.random_state = random_state
 
     def fit(self, X, y=None, sample_weight=None):
-        """Applies the Bornhuetter-Ferguson technique to triangle **X**
+        """Applies the Benktander technique to triangle **X**
 
         Parameters
         ----------
@@ -60,7 +60,7 @@ def fit(self, X, y=None, sample_weight=None):
         return self
 
     def predict(self, X, sample_weight=None):
-        """Predicts the Bornhuetter-Ferguson ultimate on a new triangle **X**
+        """Predicts the Benktander ultimate on a new triangle **X**
 
         Parameters
         ----------

chainladder/methods/tests/test_predict.py

@@ -111,7 +111,7 @@ def test_misaligned_index2(clrd):
     b = bcc.predict(clrd.iloc[150], sample_weight=w.iloc[150]).ultimate_.sum().sum()
     assert abs(a - b) < 1e-5
 
-def align_cdfs():
+def test_align_cdfs():
     ld = cl.load_sample('raa').latest_diagonal*0+40000
     model = cl.BornhuetterFerguson().fit(cl.load_sample('raa'), sample_weight=ld)
     a = model.ultimate_.iloc[..., :4, :] 
@@ -127,6 +127,6 @@ def align_cdfs():
     assert a == b
 
 
-def check_val_tri_cl(raa):
+def test_check_val_tri_cl(raa):
     model = cl.Chainladder().fit(raa.dev_to_val())
     assert model.predict(raa.latest_diagonal).ultimate_ == model.ultimate_

ci/environment-latest.yaml

@@ -6,6 +6,7 @@ channels:
 
 dependencies:
   - conda-forge::pandas>=2.0
+  - polars
   - scikit-learn>=0.23
   - sparse>=0.9
   - numba

ci/environment-test.yaml

@@ -7,6 +7,7 @@ channels:
 
 dependencies:
   - pandas>=0.23
+  - polars>=0.16.2
   - scikit-learn>=0.23
   - sparse>=0.9
   - numba

conftest.py

@@ -66,7 +66,7 @@ def prism(request):
 @pytest.fixture
 def prism_dense(request):
     if request.param == "sparse_only_run":
-        cl.options.set_option('ARRAY_BACKEND', 'numpy')
+        cl.options.set_option('ARRAY_BACKEND', 'sparse')
     else:
         cl.options.set_option('ARRAY_BACKEND', 'numpy')
     return cl.load_sample('prism').sum()

docs/gallery/plot_benktander.ipynb

@@ -27,7 +27,7 @@
    "metadata": {},
    "source": [
     "This example demonstrates the relationship between the Chainladder and\n",
-    "BornhuetterFerguson methods by way fo the Benktander model. Each is a\n",
+    "BornhuetterFerguson methods by way of the Benktander model. Each is a\n",
     "special case of the Benktander model where ``n_iters = 1`` for BornhuetterFerguson\n",
     "and as ``n_iters`` approaches infinity yields the chainladder.  As ``n_iters``\n",
     "increases the apriori selection becomes less relevant regardless of initial\n",

docs/user_guide/adjustments.ipynb

@@ -558,7 +558,7 @@
     "vertical line estimates as well as the more common effective date estimates.\n",
     "\n",
     "```{eval-rst}\n",
-    ".. image:: images/onlevel.PNG\n",
+    ".. image:: ../images/onlevel.PNG\n",
     "```\n",
     "\n",
     "### Rate History\n",

docs/user_guide/development.ipynb

@@ -5459,7 +5459,7 @@
     "which assumes no ``valuation`` covariate, the PTF framework allows for this.\n",
     "\n",
     "```{eval-rst}\n",
-    ".. figure:: images/prob_trend_family.PNG\n",
+    ".. figure:: ../images/prob_trend_family.PNG\n",
     "   :align: center\n",
     "   :scale: 40%\n",
     "```\n",

docs/user_guide/workflow.ipynb

@@ -1653,6 +1653,8 @@
     "Using `GridSearch` for scenario testing is entirely optional.  You can write\n",
     "your own looping mechanisms to achieve the same result.  For example:\n",
     "\n",
+    "<!-- TODO: what is supposed to be showing here? -->\n",
+    "\n",
     "```{eval-rst}\n",
     ".. figure:: /auto_examples/images/sphx_glr_plot_capecod_001.png\n",
     "   :target: ../auto_examples/plot_capecod.html\n",

environment-dev.yaml

@@ -14,6 +14,7 @@ dependencies:
   - ipykernel
 
   - pandas
+  - polars
   - scikit-learn
   - sparse
   - numba

environment-docs.yaml

@@ -11,6 +11,7 @@ dependencies:
   - git
 
   - pandas
+  - polars
   - scikit-learn
   - sparse
   - numba