explainiverse 0.1.1a1__py3-none-any.whl → 0.2.0__py3-none-any.whl

explainiverse/__init__.py

@@ -1 +1,45 @@
-# Explainiverse Init
+# src/explainiverse/__init__.py
+"""
+Explainiverse - A unified, extensible explainability framework.
+
+Supports multiple XAI methods including LIME, SHAP, Anchors, Counterfactuals,
+Permutation Importance, PDP, ALE, and SAGE through a consistent interface.
+
+Quick Start:
+    from explainiverse import default_registry
+    
+    # List available explainers
+    print(default_registry.list_explainers())
+    
+    # Create an explainer
+    explainer = default_registry.create("lime", model=adapter, training_data=X, ...)
+    explanation = explainer.explain(instance)
+"""
+
+from explainiverse.core.explainer import BaseExplainer
+from explainiverse.core.explanation import Explanation
+from explainiverse.core.registry import (
+    ExplainerRegistry,
+    ExplainerMeta,
+    default_registry,
+    get_default_registry,
+)
+from explainiverse.adapters.sklearn_adapter import SklearnAdapter
+from explainiverse.engine.suite import ExplanationSuite
+
+__version__ = "0.2.0"
+
+__all__ = [
+    # Core
+    "BaseExplainer",
+    "Explanation",
+    # Registry
+    "ExplainerRegistry",
+    "ExplainerMeta",
+    "default_registry",
+    "get_default_registry",
+    # Adapters
+    "SklearnAdapter",
+    # Engine
+    "ExplanationSuite",
+]

explainiverse/adapters/__init__.py

@@ -0,0 +1,9 @@
+# src/explainiverse/adapters/__init__.py
+"""
+Model adapters - wrappers that provide a consistent interface for different ML frameworks.
+"""
+
+from explainiverse.adapters.base_adapter import BaseModelAdapter
+from explainiverse.adapters.sklearn_adapter import SklearnAdapter
+
+__all__ = ["BaseModelAdapter", "SklearnAdapter"]

explainiverse/adapters/base_adapter.py

@@ -1,25 +1,25 @@
-# src/explainiverse/adapters/base_adapter.py
-
-from abc import ABC, abstractmethod
-
-class BaseModelAdapter(ABC):
-    """
-    Abstract base class for all model adapters.
-    """
-
-    def __init__(self, model, feature_names=None):
-        self.model = model
-        self.feature_names = feature_names
-
-    @abstractmethod
-    def predict(self, data):
-        """
-        Returns prediction probabilities or outputs in a standard format.
-
-        Args:
-            data: Input data (single instance or batch).
-
-        Returns:
-            List or NumPy array of prediction scores.
-        """
-        pass
+# src/explainiverse/adapters/base_adapter.py
+
+from abc import ABC, abstractmethod
+
+class BaseModelAdapter(ABC):
+    """
+    Abstract base class for all model adapters.
+    """
+
+    def __init__(self, model, feature_names=None):
+        self.model = model
+        self.feature_names = feature_names
+
+    @abstractmethod
+    def predict(self, data):
+        """
+        Returns prediction probabilities or outputs in a standard format.
+
+        Args:
+            data: Input data (single instance or batch).
+
+        Returns:
+            List or NumPy array of prediction scores.
+        """
+        pass

explainiverse/adapters/sklearn_adapter.py

@@ -1,32 +1,32 @@
-# src/explainiverse/adapters/sklearn_adapter.py
-
-import numpy as np
-from .base_adapter import BaseModelAdapter
-
-class SklearnAdapter(BaseModelAdapter):
-    """
-    Adapter for Scikit-learn classifiers.
-    """
-
-    def __init__(self, model, feature_names=None, class_names=None):
-        super().__init__(model, feature_names)
-        self.class_names = class_names
-
-    def predict(self, data: np.ndarray) -> np.ndarray:
-        """
-        Returns prediction probabilities.
-
-        Args:
-            data: A 2D numpy array of inputs.
-
-        Returns:
-            Array of shape (n_samples, n_classes).
-        """
-        if hasattr(self.model, "predict_proba"):
-            return self.model.predict_proba(data)
-        else:
-            preds = self.model.predict(data)
-            if self.class_names:
-                return np.eye(len(self.class_names))[preds]
-            else:
-                return preds.reshape(-1, 1)  # regression: raw outputs
+# src/explainiverse/adapters/sklearn_adapter.py
+
+import numpy as np
+from .base_adapter import BaseModelAdapter
+
+class SklearnAdapter(BaseModelAdapter):
+    """
+    Adapter for Scikit-learn classifiers.
+    """
+
+    def __init__(self, model, feature_names=None, class_names=None):
+        super().__init__(model, feature_names)
+        self.class_names = class_names
+
+    def predict(self, data: np.ndarray) -> np.ndarray:
+        """
+        Returns prediction probabilities.
+
+        Args:
+            data: A 2D numpy array of inputs.
+
+        Returns:
+            Array of shape (n_samples, n_classes).
+        """
+        if hasattr(self.model, "predict_proba"):
+            return self.model.predict_proba(data)
+        else:
+            preds = self.model.predict(data)
+            if self.class_names:
+                return np.eye(len(self.class_names))[preds]
+            else:
+                return preds.reshape(-1, 1)  # regression: raw outputs

explainiverse/core/__init__.py

@@ -0,0 +1,22 @@
+# src/explainiverse/core/__init__.py
+"""
+Explainiverse core components.
+"""
+
+from explainiverse.core.explainer import BaseExplainer
+from explainiverse.core.explanation import Explanation
+from explainiverse.core.registry import (
+    ExplainerRegistry,
+    ExplainerMeta,
+    default_registry,
+    get_default_registry,
+)
+
+__all__ = [
+    "BaseExplainer",
+    "Explanation",
+    "ExplainerRegistry",
+    "ExplainerMeta",
+    "default_registry",
+    "get_default_registry",
+]

explainiverse/core/explainer.py

@@ -1,31 +1,31 @@
-# src/explainiverse/core/explainer.py
-
-from abc import ABC, abstractmethod
-
-class BaseExplainer(ABC):
-    """
-    Abstract base class for all explainers in Explainiverse.
-    """
-
-    def __init__(self, model):
-        """
-        Initialize with a model adapter or raw model.
-
-        Args:
-            model: A wrapped ML model with a standardized `predict` method.
-        """
-        self.model = model
-
-    @abstractmethod
-    def explain(self, instance, **kwargs):
-        """
-        Generate an explanation for a single input instance.
-
-        Args:
-            instance: The input to explain (e.g., feature vector, image, text).
-            **kwargs: Optional method-specific parameters.
-
-        Returns:
-            An Explanation object or dict.
-        """
-        pass
+# src/explainiverse/core/explainer.py
+
+from abc import ABC, abstractmethod
+
+class BaseExplainer(ABC):
+    """
+    Abstract base class for all explainers in Explainiverse.
+    """
+
+    def __init__(self, model):
+        """
+        Initialize with a model adapter or raw model.
+
+        Args:
+            model: A wrapped ML model with a standardized `predict` method.
+        """
+        self.model = model
+
+    @abstractmethod
+    def explain(self, instance, **kwargs):
+        """
+        Generate an explanation for a single input instance.
+
+        Args:
+            instance: The input to explain (e.g., feature vector, image, text).
+            **kwargs: Optional method-specific parameters.
+
+        Returns:
+            An Explanation object or dict.
+        """
+        pass

explainiverse/core/explanation.py

@@ -1,24 +1,24 @@
-# src/explainiverse/core/explanation.py
-
-class Explanation:
-    """
-    Unified container for explanation results.
-    """
-
-    def __init__(self, explainer_name: str, target_class: str, explanation_data: dict):
-        self.explainer_name = explainer_name
-        self.target_class = target_class
-        self.explanation_data = explanation_data  # e.g., {'feature_attributions': {...}}
-
-    def __repr__(self):
-        return (f"Explanation(explainer='{self.explainer_name}', "
-                f"target='{self.target_class}', "
-                f"keys={list(self.explanation_data.keys())})")
-
-    def plot(self, type='bar'):
-        """
-        Visualizes the explanation.
-        This will later integrate with a proper visualization backend.
-        """
-        print(f"[plot: {type}] Plotting explanation for {self.target_class} "
-              f"from {self.explainer_name}.")
+# src/explainiverse/core/explanation.py
+
+class Explanation:
+    """
+    Unified container for explanation results.
+    """
+
+    def __init__(self, explainer_name: str, target_class: str, explanation_data: dict):
+        self.explainer_name = explainer_name
+        self.target_class = target_class
+        self.explanation_data = explanation_data  # e.g., {'feature_attributions': {...}}
+
+    def __repr__(self):
+        return (f"Explanation(explainer='{self.explainer_name}', "
+                f"target='{self.target_class}', "
+                f"keys={list(self.explanation_data.keys())})")
+
+    def plot(self, type='bar'):
+        """
+        Visualizes the explanation.
+        This will later integrate with a proper visualization backend.
+        """
+        print(f"[plot: {type}] Plotting explanation for {self.target_class} "
+              f"from {self.explainer_name}.")