Merge branch '21637-add-thresholds-auto-ablation' of github.com:howso…

…ai/howso-engine-py into 21637-add-thresholds-auto-ablation

howso/client/base.py

@@ -1848,7 +1848,7 @@ def react(  # noqa: C901
             If set to True, will scale influence weights by each case's
             `weight_feature` weight. If unspecified, case weights
             will be used if the Trainee has them.
-        case_indices : Iterable of Sequence[Union[str, int]], defaults to None
+        case_indices : Iterable of Sequence[str | int], defaults to None
             An Iterable of Sequences, of session id and index, where
             index is the original 0-based index of the case as it was trained
             into the session. If this case does not exist, discriminative react
@@ -1956,7 +1956,7 @@ def react(  # noqa: C901
                 action -> pandas.DataFrame
                     A data frame of action values.
 
-                details -> Dict or List
+                details -> dict or list
                     An aggregated list of any requested details.
 
         Raises
@@ -2695,7 +2695,7 @@ def react_series(  # noqa: C901
                 action -> pandas.DataFrame
                     A data frame of action values.
 
-                details -> Dict or List
+                details -> dict or list
                     An aggregated list of any requested details.
 
         Raises
@@ -4062,10 +4062,10 @@ def set_auto_ablation_params(
         residual_prediction_features : Optional[List[str]], optional
             For each of the features specified, will ablate a case if
             abs(prediction - case value) / prediction <= feature residual.
-        tolerance_prediction_threshold_map : Optional[Dict[str, Tuple[float, float]]], optional
+        tolerance_prediction_threshold_map : Optional[dict[str, tuple[float, float]]], optional
             For each of the features specified, will ablate a case if the prediction >= (case value - MIN)
             and the prediction <= (case value + MAX).
-        relative_prediction_threshold_map : Optional[Dict[str, float]], optional
+        relative_prediction_threshold_map : Optional[dict[str, float]], optional
             For each of the features specified, will ablate a case if
             abs(prediction - case value) / prediction <= relative threshold
         conviction_lower_threshold : Optional[float], optional

howso/client/feature_flags.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import typing as t
 import warnings
 
@@ -13,9 +15,9 @@ class FeatureFlags:
     """
 
     # Define obsolete flags here to raise a warning when defined
-    _obsolete_flags: t.Union[t.Set[str], None] = None
+    _obsolete_flags: set[str] | None = None
 
-    def __init__(self, flags: t.Optional[t.Dict[str, t.Any]]):
+    def __init__(self, flags: t.Optional[dict[str, t.Any]]):
         self._store = dict()
         if flags is not None:
             obsolete = set()
@@ -55,7 +57,7 @@ def parse_flag(cls, flag: str) -> str:
         """Parse the flag name."""
         return flag.replace('-', '_').lower()
 
-    def __iter__(self) -> t.Generator[t.Tuple[str, bool], None, None]:
+    def __iter__(self) -> t.Generator[tuple[str, bool], None, None]:
         """Iterate over flags."""
         return ((key, value) for key, value in self._store.items())
 

howso/client/pandas/client.py

@@ -1,7 +1,7 @@
 from __future__ import annotations
 
-from collections.abc import Iterable
-from typing import Optional
+from collections.abc import Collection
+import typing as t
 
 import pandas as pd
 from pandas import DataFrame, Index
@@ -88,7 +88,7 @@ def get_extreme_cases(
         trainee_id: str,
         num: int,
         sort_feature: str,
-        features: Optional[Iterable[str]] = None
+        features: t.Optional[Collection[str]] = None
     ) -> DataFrame:
         """
         Base: :func:`howso.client.AbstractHowsoClient.get_extreme_cases`.

howso/client/schemas/reaction.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 from collections import abc
 from functools import singledispatchmethod
 from pprint import pformat
@@ -29,11 +31,11 @@ class Reaction(abc.MutableMapping):
 
     Parameters
     ----------
-    action : Union[pandas.DataFrame, list, dict], default None
+    action : pandas.DataFrame or list or dict, default None
         (Optional) A DataFrame with columns representing the requested
         features of ``react`` or ``react_series`` cases.
 
-    details : List or None
+    details : list or None
         (Optional) The details of results from ``react`` or ``react_series``
         when providing a ``details`` parameter.
     """
@@ -60,8 +62,8 @@ class Reaction(abc.MutableMapping):
     }
 
     def __init__(self,
-                 action: t.Optional[t.Union[pd.DataFrame, list, dict]] = None,
-                 details: t.Optional[t.MutableMapping[str, t.Any]] = None
+                 action: t.Optional[pd.DataFrame | list | dict] = None,
+                 details: t.Optional[abc.MutableMapping[str, t.Any]] = None
                  ):
         """Initialize the dictionary with the allowed keys."""
         self._data = {
@@ -79,7 +81,7 @@ def __init__(self,
 
         self._reorganized_details = None
 
-    def _validate_key(self, key) -> str:
+    def _validate_key(self, key: str) -> str:
         """
         Raise KeyError if key is not one of the allowed keys.
 
@@ -115,18 +117,18 @@ def _validate_key(self, key) -> str:
 
         return key
 
-    def __getitem__(self, key):
+    def __getitem__(self, key: str):
         """Get an item by key if the key is allowed."""
         key = self._validate_key(key)
         return self._data[key]
 
-    def __setitem__(self, key, value):
+    def __setitem__(self, key: str, value: t.Any):
         """Set an item by key if the key is allowed."""
         key = self._validate_key(key)
         self._reorganized_details = None
         self._data[key] = value
 
-    def __delitem__(self, key):
+    def __delitem__(self, key: str):
         """Delete an item by key if the key is allowed."""
         key = self._validate_key(key)
         self._reorganized_details = None
@@ -148,7 +150,7 @@ def __repr__(self) -> str:
 
     @singledispatchmethod
     def add_reaction(self, action: pd.DataFrame,
-                     details: t.MutableMapping[str, t.Any]):
+                     details: abc.MutableMapping[str, t.Any]):
         """
         Add more data to the instance.
 
@@ -201,18 +203,18 @@ def add_reaction(self, action: pd.DataFrame,
         self._reorganized_details = None
 
     @add_reaction.register
-    def _(self, action: dict, details: t.MutableMapping[str, t.Any]):
-        """Add Dict[List, Dict] to Reaction."""
+    def _(self, action: dict, details: abc.MutableMapping[str, t.Any]):
+        """Add dict[list, dict] to Reaction."""
         action_df = pd.DataFrame.from_dict(action)
         return self.add_reaction(action_df, details)
 
     @add_reaction.register
-    def _(self, action: list, details: t.MutableMapping[str, t.Any]):
-        """Add list[Dict] to Reaction."""
+    def _(self, action: list, details: abc.MutableMapping[str, t.Any]):
+        """Add list[dict] to Reaction."""
         action_df = pd.DataFrame(action)
         return self.add_reaction(action_df, details)
 
-    def gen_cases(self) -> t.Generator[t.Dict, None, None]:
+    def gen_cases(self) -> t.Generator[dict, None, None]:
         """
         Yield dict containing DetailedCase items for a single case.
 
@@ -240,8 +242,7 @@ def reorganized_details(self):
         return self._reorganized_details
 
     @classmethod
-    def _reorganize_details(cls, details: t.MutableMapping[str, t.List]
-                            ) -> t.List[t.Dict]:
+    def _reorganize_details(cls, details: abc.MutableMapping[str, list]) -> list[dict]:
         """
         Re-organize `details` to be a list of dicts. One dict per case.
 
@@ -261,11 +262,15 @@ def _reorganize_details(cls, details: t.MutableMapping[str, t.List]
                 {k1: v1m, k2: v2m, ... kn: vnm}
             ]
 
-        Parameters:
-            details : Dict of Lists
+        Parameters
+        ----------
+        details : dict of list
+            The reaction details.
 
-        Returns:
-            List of Dicts, one Dict per case
+        Returns
+        -------
+        List of dicts
+            One dict per case.
         """
         if isinstance(details, list):
             return details
@@ -282,7 +287,7 @@ def _reorganize_details(cls, details: t.MutableMapping[str, t.List]
             k: v for k, v in details.items()
             if k in cls.KNOWN_KEYS and v
         }
-        # Transform Dict[List] -> List[Dict]
+        # Transform dict[list] -> list[dict]
         per_case_details = [
             dict(zip([key for key in cleaned_details.keys()], values))
             for values in zip(*cleaned_details.values())

howso/direct/_utilities.py

@@ -1,10 +1,11 @@
+from __future__ import annotations
+
 from importlib import metadata
 from pathlib import Path
 import sysconfig
-from typing import Union
 
 
-def get_file_in_distribution(file_path) -> Union[Path, None]:
+def get_file_in_distribution(file_path: str) -> Path | None:
     """
     Locate the LICENSE.txt file in the distribution of this package.
 
@@ -20,6 +21,8 @@ def get_file_in_distribution(file_path) -> Union[Path, None]:
     """
     purelib_path = sysconfig.get_path('purelib')
     dist = metadata.distribution('howso-engine')
+    if dist.files is None:
+        raise AssertionError("The package howso-engine is not installed correctly, please reinstall.")
     for fp in dist.files:
         if fp.name == file_path:
             return Path(purelib_path, fp)

howso/engine/trainee.py

@@ -312,7 +312,7 @@ def needs_analyze(self) -> bool:
         return self._needs_analyze
 
     @property
-    def calculated_matrices(self) -> t.Optional[dict[str, DataFrame]]:
+    def calculated_matrices(self) -> dict[str, DataFrame] | None:
         """
         The calculated matrices.
 
@@ -2651,7 +2651,7 @@ def react_group(
         use_case_weights: t.Optional[bool] = None,
         features: t.Optional[Collection[str]] = None,
         weight_feature: t.Optional[str] = None,
-    ) -> DataFrame | dict:
+    ) -> DataFrame:
         """
         Computes specified data for a **set** of cases.
 
@@ -2702,23 +2702,26 @@ def react_group(
 
         Returns
         -------
-        DataFrame or dict
+        DataFrame
             The conviction of grouped cases.
         """
-        return self.client.react_group(
-            trainee_id=self.id,
-            new_cases=new_cases,
-            features=features,
-            familiarity_conviction_addition=familiarity_conviction_addition,
-            familiarity_conviction_removal=familiarity_conviction_removal,
-            kl_divergence_addition=kl_divergence_addition,
-            kl_divergence_removal=kl_divergence_removal,
-            p_value_of_addition=p_value_of_addition,
-            p_value_of_removal=p_value_of_removal,
-            distance_contributions=distance_contributions,
-            use_case_weights=use_case_weights,
-            weight_feature=weight_feature,
-        )
+        if isinstance(self.client, HowsoPandasClientMixin):
+            return self.client.react_group(
+                trainee_id=self.id,
+                new_cases=new_cases,
+                features=features,
+                familiarity_conviction_addition=familiarity_conviction_addition,
+                familiarity_conviction_removal=familiarity_conviction_removal,
+                kl_divergence_addition=kl_divergence_addition,
+                kl_divergence_removal=kl_divergence_removal,
+                p_value_of_addition=p_value_of_addition,
+                p_value_of_removal=p_value_of_removal,
+                distance_contributions=distance_contributions,
+                use_case_weights=use_case_weights,
+                weight_feature=weight_feature,
+            )
+        else:
+            raise AssertionError("Client must have the 'react_group' method.")
 
     def get_feature_conviction(
         self,
@@ -2729,7 +2732,7 @@ def get_feature_conviction(
         action_features: t.Optional[Collection[str]] = None,
         features: t.Optional[Collection[str]] = None,
         weight_feature: t.Optional[str] = None,
-    ) -> DataFrame | dict:
+    ) -> DataFrame:
         """
         Get familiarity conviction for features in the model.
 
@@ -2761,19 +2764,22 @@ def get_feature_conviction(
 
         Returns
         -------
-        DataFrame or dict
+        DataFrame
             A DataFrame containing the familiarity conviction rows to feature
             columns.
         """
-        return self.client.get_feature_conviction(
-            trainee_id=self.id,
-            action_features=action_features,
-            familiarity_conviction_addition=familiarity_conviction_addition,
-            familiarity_conviction_removal=familiarity_conviction_removal,
-            features=features,
-            use_case_weights=use_case_weights,
-            weight_feature=weight_feature,
-        )
+        if isinstance(self.client, HowsoPandasClientMixin):
+            return self.client.get_feature_conviction(
+                trainee_id=self.id,
+                action_features=action_features,
+                familiarity_conviction_addition=familiarity_conviction_addition,
+                familiarity_conviction_removal=familiarity_conviction_removal,
+                features=features,
+                use_case_weights=use_case_weights,
+                weight_feature=weight_feature,
+            )
+        else:
+            raise AssertionError("Client must have the 'get_feature_conviction' method.")
 
     def get_marginal_stats(
         self, *,