prep 0.8.2

README.rst

@@ -60,6 +60,8 @@ Available Estimators
 +------------------------------+------------------+-------------------------+-----------------------+-----------------------+
 | `DevelopmentML`_             |                  |                         |                       |                       |
 +------------------------------+------------------+-------------------------+-----------------------+-----------------------+
+| `BarnettZehnwirth`_          |                  |                         |                       |                       |
++------------------------------+------------------+-------------------------+-----------------------+-----------------------+
 
 Documentation
 -------------
@@ -91,6 +93,7 @@ code documentation.
 .. _CaseOutstanding: https://chainladder-python.readthedocs.io/en/latest/modules/development.html#caseoutstanding
 .. _TweedieGLM: https://chainladder-python.readthedocs.io/en/latest/modules/development.html#tweedieglm
 .. _DevelopmentML: https://chainladder-python.readthedocs.io/en/latest/modules/development.html#developmentml
+.. _BarnettZehnwirth: https://chainladder-python.readthedocs.io/en/latest/modules/development.html#barnettzehnwirth
 .. _Documentation: https://chainladder-python.readthedocs.io/en/latest/
 
 Getting Started Tutorials

chainladder/core/triangle.py

@@ -331,21 +331,23 @@ def dev_to_val(self, inplace=False):
             else:
                 return self.copy()
         is_cumulative = self.is_cumulative
-        if self.is_ultimate:
+        if self.is_full:
             if is_cumulative:
                 obj = self.cum_to_incr(inplace=inplace)
             else:
                 obj = self.copy()
-            ultimate = obj.iloc[..., -1:]
-            obj = obj.iloc[..., :-1]
+            if self.is_ultimate:
+                ultimate = obj.iloc[..., -1:]
+                obj = obj.iloc[..., :-1]
         else:
             obj = self
         obj = obj._val_dev(1, inplace)
         ddims = obj.valuation[obj.valuation <= obj.valuation_date]
         obj.ddims = ddims.drop_duplicates().sort_values()
-        if self.is_ultimate:
-            ultimate.ddims = pd.DatetimeIndex(ultimate.valuation[0:1])
-            obj = concat((obj, ultimate), -1)
+        if self.is_full:
+            if self.is_ultimate:
+                ultimate.ddims = pd.DatetimeIndex(ultimate.valuation[0:1])
+                obj = concat((obj, ultimate), -1)
             if is_cumulative:
                 obj = obj.incr_to_cum(inplace=inplace)
         return obj
@@ -427,6 +429,10 @@ def grain(self, grain="", trailing=False, inplace=False):
             )
         if valid["M"].index(ograin_new) > valid["M"].index(dgrain_new):
             raise ValueError("Origin grain must be coarser than development grain")
+        if self.is_full and not self.is_ultimate and not self.is_val_tri:
+            warnings.warn('Triangle includes extraneous development lags')
+        else:
+            d_limit = None
         obj = self.dev_to_val()
         if ograin_new != ograin_old:
             if trailing:

chainladder/development/barnzehn.py

@@ -38,7 +38,7 @@ def fit(self, X, y=None, sample_weight=None):
         response = X.columns[0] if not self.response else self.response
         self.model_ = DevelopmentML(Pipeline(steps=[
             ('design_matrix', PatsyFormula(self.formula)),
-            ('model', LinearRegression(fit_intercept=True))]),
+            ('model', LinearRegression(fit_intercept=False))]),
                     y_ml=response, fit_incrementals=False).fit(tri)
         resid = tri - self.model_.triangle_ml_[
             self.model_.triangle_ml_.valuation <= tri.valuation_date]

chainladder/development/base.py

@@ -297,8 +297,9 @@ def fit(self, X, y=None, sample_weight=None):
         self.sigma_ = self._param_property(X, params, 1)
         self.std_err_ = self._param_property(X, params, 2)
 
-        resid = -X.iloc[..., :-1]*self.ldf_.values+X.iloc[..., 1:].values
-        std = xp.sqrt((1/w)*(self.sigma_**2).values)
+        resid = -X.iloc[..., :-1] * self.ldf_.values + X.iloc[..., 1:].values
+
+        std = xp.sqrt((1/num_to_nan(w))*(self.sigma_**2).values)
         resid = resid/std
         self.std_residuals_ = resid[resid.valuation < X.valuation_date]
         return self

chainladder/development/learning.py

@@ -204,7 +204,8 @@ def transform(self, X):
         X_ml = self._prep_X_ml(X)
         y_ml=self.estimator_ml.predict(X_ml)
         triangle_ml = self._get_triangle_ml(X_ml, y_ml)
-        X_new.ldf_ = triangle_ml.incr_to_cum().link_ratio
+        backend = "cupy" if X.array_backend == "cupy" else "numpy"
+        X_new.ldf_ = triangle_ml.incr_to_cum().link_ratio.set_backend(backend)
         X_new.ldf_.valuation_date = pd.to_datetime(ULT_VAL)
         X_new._set_slicers()
         return X_new

chainladder/development/outstanding.py

@@ -66,6 +66,7 @@ def fit(self, X, y=None, sample_weight=None):
         self : object
             Returns the instance itself.
         """
+        backend = "cupy" if X.array_backend == "cupy" else "numpy"
         self.X_ = X.copy()
         paid_tri = self.X_[self.paid_to_incurred[0]]
         incurred_tri = self.X_[self.paid_to_incurred[1]]
@@ -104,7 +105,8 @@ def fit(self, X, y=None, sample_weight=None):
         dev.ddims = self.X_.link_ratio.ddims
         dev.is_pattern=True
         dev.is_cumulative=True
-        self.ldf_ = dev.cum_to_incr()
+
+        self.ldf_ = dev.cum_to_incr().set_backend(backend)
         return self
 
     @property

chainladder/methods/mack.py

@@ -141,7 +141,7 @@ def _mack_recursion(self, est, X=None):
         extend = X.ldf_.shape[-1] - X.shape[-1] + 1
         ldf = X.ldf_.values[..., : len(X.ddims) - 1]
         tail = X.cdf_.values[..., -extend : -extend + 1]
-        ldf = X.ldf_.get_array_module().concatenate((ldf, tail), -1)
+        ldf = xp.array(X.ldf_.get_array_module().concatenate((ldf, tail), -1))
         # Recursive Mack Formula
         for i in range(t1_t.shape[-1]):
             t1 = t1_t[..., i : i + 1] ** 2

chainladder/utils/weighted_regression.py

@@ -8,8 +8,8 @@
 
 class WeightedRegression(BaseEstimator):
     """ Helper class that fits a system of regression equations
-        simultaneously on a multi-dimensional array.  Look into
-        SUR as a replacement.
+        as a closed-form solution.  This greatly speeds up
+        the implementation of the Mack stochastic properties.
     """
 
     def __init__(self, axis=None, thru_orig=False, xp=None):