probablity summing issue is suppressed because of sigmoid output of t…

…he models
scikit-learn-contrib · Dec 12, 2024 · ce2d960 · ce2d960
1 parent 0a25434
commit ce2d960
Show file tree

Hide file tree

Showing 8 changed files with 439 additions and 95 deletions.
diff --git a/mapie/conformity_scores/sets/aps.py b/mapie/conformity_scores/sets/aps.py
@@ -78,7 +78,7 @@ def get_predictions(
             Array of predictions.
         """
         y_pred_proba = estimator.predict(X, agg_scores)
-        y_pred_proba = check_proba_normalized(y_pred_proba, axis=1)
+        # y_pred_proba = check_proba_normalized(y_pred_proba, axis=1)
         if agg_scores != "crossval":
             y_pred_proba = np.repeat(
                 y_pred_proba[:, :, np.newaxis], len(alpha_np), axis=2
@@ -161,6 +161,7 @@ def get_conformity_scores(
         y_proba_true = np.take_along_axis(
             y_pred, y_enc.reshape(-1, 1), axis=1
         )
+
         random_state = check_random_state(self.random_state)
         u = random_state.uniform(size=len(y_pred)).reshape(-1, 1)
         conformity_scores -= u * y_proba_true

diff --git a/mapie/conformity_scores/sets/lac.py b/mapie/conformity_scores/sets/lac.py
@@ -112,7 +112,7 @@ def get_predictions(
             Array of predictions.
         """
         y_pred_proba = estimator.predict(X, agg_scores)
-        y_pred_proba = check_proba_normalized(y_pred_proba, axis=1)
+        # y_pred_proba = check_proba_normalized(y_pred_proba, axis=1)
         if agg_scores != "crossval":
             y_pred_proba = np.repeat(
                 y_pred_proba[:, :, np.newaxis], len(alpha_np), axis=2

diff --git a/mapie/conformity_scores/sets/naive.py b/mapie/conformity_scores/sets/naive.py
@@ -85,7 +85,7 @@ def get_predictions(
             Array of predictions.
         """
         y_pred_proba = estimator.predict(X, agg_scores='mean')
-        y_pred_proba = check_proba_normalized(y_pred_proba, axis=1)
+        # y_pred_proba = check_proba_normalized(y_pred_proba, axis=1)
         y_pred_proba = np.repeat(
             y_pred_proba[:, :, np.newaxis], len(alpha_np), axis=2
         )

diff --git a/mapie/conformity_scores/sets/topk.py b/mapie/conformity_scores/sets/topk.py
@@ -110,7 +110,7 @@ def get_predictions(
             Array of predictions.
         """
         y_pred_proba = estimator.predict(X, agg_scores="mean")
-        y_pred_proba = check_proba_normalized(y_pred_proba, axis=1)
+        # y_pred_proba = check_proba_normalized(y_pred_proba, axis=1)
         y_pred_proba = np.repeat(
             y_pred_proba[:, :, np.newaxis], len(alpha_np), axis=2
         )

diff --git a/mapie/estimator/classifier.py b/mapie/estimator/classifier.py
@@ -416,7 +416,7 @@ def predict_proba_calib(
 
         if self.cv == "prefit":
             y_pred_proba = self.single_estimator_.predict_proba(X)
-            y_pred_proba = self._check_proba_normalized(y_pred_proba)
+            # y_pred_proba = self._check_proba_normalized(y_pred_proba)
         else:
             X = cast(NDArray, X)
             y_pred_proba = np.empty((len(X), self.n_classes), dtype=float)

diff --git a/notebooks/classification/Cifar10.ipynb b/notebooks/classification/Cifar10.ipynb
diff --git a/notebooks/classification/text.ipynb b/notebooks/classification/text.ipynb
@@ -0,0 +1,24 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "detectron2",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "name": "python",
+   "version": "3.12.4"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}
diff --git a/text.ipynb b/text.ipynb
@@ -0,0 +1,56 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from mapie.conformity_scores.sets.topk import TopKConformityScore"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "metadata": {},
+   "outputs": [
+    {
+     "ename": "TypeError",
+     "evalue": "TopKConformityScore.get_conformity_scores() missing 2 required positional arguments: 'y' and 'y_pred'",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
+      "\u001b[1;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
+      "Cell \u001b[1;32mIn[6], line 2\u001b[0m\n\u001b[0;32m      1\u001b[0m topK \u001b[38;5;241m=\u001b[39m TopKConformityScore()\n\u001b[1;32m----> 2\u001b[0m \u001b[43mtopK\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget_conformity_scores\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n",
+      "\u001b[1;31mTypeError\u001b[0m: TopKConformityScore.get_conformity_scores() missing 2 required positional arguments: 'y' and 'y_pred'"
+     ]
+    }
+   ],
+   "source": [
+    "topK = TopKConformityScore()\n",
+    "topK.get_conformity_scores()"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "detectron2",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.12.4"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}