chore: update security and exaplainability

src/holisticai/bias/mitigation/inprocessing/adversarial_debiasing/transformer.py

@@ -6,6 +6,7 @@
 import jax
 import jax.numpy as jnp
 import numpy as np
+import pandas as pd
 from holisticai.bias.mitigation.inprocessing.adversarial_debiasing.models import (
     ADModel,
     AdversarialModel,
@@ -20,6 +21,10 @@
 logger = logging.getLogger(__name__)
 
 
+def is_numeric(df):
+    return all(pd.api.types.is_numeric_dtype(df[col]) for col in df.columns)
+
+
 class AdversarialDebiasing(BMImp):
     """Adversarial Debiasing
 
@@ -176,6 +181,9 @@ def fit(
 
         params = self._load_data(X=X, y=y, group_a=group_a, group_b=group_b)
         x = pd.DataFrame(params["X"])
+        if not is_numeric(x):
+            raise ValueError("Adversarial Debiasing only works with numeric features.")
+
         y = pd.Series(params["y"])
         group_a = pd.Series(params["group_a"])
         group_b = pd.Series(params["group_b"])
@@ -239,6 +247,8 @@ def predict(self, X):
 
         np.ndarray: Predicted output per sample.
         """
+        if not is_numeric(X):
+            raise ValueError("Adversarial Debiasing only works with numeric features.")
         p = self.predict_proba(X)
         return np.argmax(p, axis=1).ravel()
 
@@ -260,6 +270,9 @@ def predict_proba(self, X):
 
         np.ndarray: Predicted matrix probability per sample.
         """
+        if not is_numeric(X):
+            raise ValueError("Adversarial Debiasing only works with numeric features.")
+
         proba = np.empty((X.shape[0], 2))
         proba[:, 1] = self._predict_proba(X)
         proba[:, 0] = 1.0 - proba[:, 1]
@@ -283,5 +296,8 @@ def predict_score(self, X):
 
         np.ndarray: Predicted probability per sample.
         """
+        if not is_numeric(X):
+            raise ValueError("Adversarial Debiasing only works with numeric features.")
+
         p = self._predict(X).reshape([-1])
         return p

src/holisticai/bias/mitigation/inprocessing/prejudice_remover/algorithm_utils.py

@@ -1,3 +1,4 @@
+import numpy as np
 from holisticai.bias.mitigation.inprocessing.commons import Logging
 
 
@@ -30,7 +31,7 @@ def loss(self, coef_, X, y, groups):
         loss : float
             loss function value
         """
-        coef = coef_.reshape(self.estimator.nb_group_values, self.estimator.nb_features)
+        coef = coef_.reshape(self.estimator.nb_group_values, self.estimator.nb_features).astype(np.float64)
         X = self.estimator.preprocessing_data(X)
         sigma = self.estimator.sigmoid(X=X, groups=groups, coef=coef)
         loss = self.loss_fn(y=y, sigma=sigma, groups=groups, coef=coef)
@@ -53,7 +54,7 @@ def grad_loss(self, coef_, X, y, groups):
             first derivative of loss function
         """
 
-        coef = coef_.reshape(self.estimator.nb_group_values, self.estimator.nb_features)
+        coef = coef_.reshape(self.estimator.nb_group_values, self.estimator.nb_features).astype(np.float64)
         X = self.estimator.preprocessing_data(X)
         sigma = self.estimator.sigmoid(X=X, groups=groups, coef=coef)
         return self.loss_fn.gradient(X=X, y=y, sigma=sigma, groups=groups, coef=coef)

src/holisticai/datasets/_dataset.py

@@ -1,6 +1,6 @@
 from __future__ import annotations
 
-from typing import TYPE_CHECKING
+from typing import TYPE_CHECKING, Union
 
 import pandas as pd
 from sklearn.model_selection import train_test_split
@@ -11,6 +11,7 @@
     from collections.abc import Iterable
 
 import numpy as np
+from numpy.random import RandomState
 
 
 class DatasetDict(dict):
@@ -170,7 +171,7 @@ def __init__(self, dataset, batch_size, dtype):
     def batched(self):
         def batch_generator(batch_size):
             for i in range(self.num_batches):
-                batch = Dataset(data=self.dataset.data.iloc[i * batch_size : (i + 1) * batch_size])
+                batch = Dataset(self.dataset.data.iloc[i * batch_size : (i + 1) * batch_size])
                 yield batch
 
         if self.dtype == "jax":
@@ -242,8 +243,8 @@ def __update_metadata(self):
         features_counts = features_values.value_counts()
         self.features_is_series = {key: (value == 1) for key, value in features_counts.items()}
 
-    def __init__(self, data: pd.DataFrame | None = None, **kargs):
-        if data is None:
+    def __init__(self, _data: pd.DataFrame | None = None, **kargs):
+        if _data is None:
             self.data = {}
             for name, value in kargs.items():
                 if isinstance(value, pd.DataFrame):
@@ -257,7 +258,7 @@ def __init__(self, data: pd.DataFrame | None = None, **kargs):
             self.data.columns = self.data.columns.set_names(["features", "subfeatures"])
             self.data.reset_index(drop=True)
         else:
-            self.data = data.reset_index(drop=True)
+            self.data = _data.reset_index(drop=True)
         self.__update_metadata()
         self.random_state = np.random.RandomState()
 
@@ -419,7 +420,7 @@ def apply_fn_to_multilevel_df(df, fn):
     return result_df
 
 
-def sample_n(group, n, random_state=None):
+def sample_n(group: pd.DataFrame, n: int, random_state: Union[RandomState, None] = None) -> pd.DataFrame:
     if len(group) < n:
         return group
     return group.sample(n=n, replace=False, random_state=random_state)