mlquantify 0.1.14__py3-none-any.whl → 0.1.16__py3-none-any.whl

mlquantify/adjust_counting/_adjustment.py

@@ -2,6 +2,7 @@ import numpy as np
 from abc import abstractmethod
 from scipy.optimize import minimize
 import warnings
+from sklearn.metrics import confusion_matrix
 
 from mlquantify.adjust_counting._base import BaseAdjustCount
 from mlquantify.adjust_counting._counting import CC, PCC
@@ -203,13 +204,13 @@ class MatrixAdjustment(BaseAdjustCount):
         self.CM = np.zeros((n_class, n_class))
 
         if self.solver == 'optim':
-            priors = np.array(list(CC().aggregate(train_y_pred).values()))
+            priors = np.array(list(CC().aggregate(train_y_pred, train_y_values).values()))
             self.CM = self._compute_confusion_matrix(train_y_pred, train_y_values, priors)
-            prevs_estim = self._get_estimations(predictions > priors)
+            prevs_estim = self._get_estimations(predictions > priors, train_y_values)
             prevalence = self._solve_optimization(prevs_estim, priors)
         else:
-            self.CM = self._compute_confusion_matrix(train_y_pred)
-            prevs_estim = self._get_estimations(predictions)
+            self.CM = self._compute_confusion_matrix(train_y_pred, train_y_values)
+            prevs_estim = self._get_estimations(predictions, train_y_values)
             prevalence = self._solve_linear(prevs_estim)
         
         return prevalence
@@ -260,11 +261,11 @@ class MatrixAdjustment(BaseAdjustCount):
         result = minimize(objective, init, constraints=constraints, bounds=bounds)
         return result.x if result.success else priors
 
-    def _get_estimations(self, predictions):
+    def _get_estimations(self, predictions, train_y_values):
         """Return prevalence estimates using CC (crisp) or PCC (probabilistic)."""
         if uses_soft_predictions(self):
             return np.array(list(PCC().aggregate(predictions).values()))
-        return np.array(list(CC().aggregate(predictions).values()))
+        return np.array(list(CC().aggregate(predictions, train_y_values).values()))
 
     @abstractmethod
     def _compute_confusion_matrix(self, predictions, *args):
@@ -389,8 +390,11 @@ class GAC(CrispLearnerQMixin, MatrixAdjustment):
     def __init__(self, learner=None):
         super().__init__(learner=learner, solver='linear')
     
-    def _compute_confusion_matrix(self, predictions):
-        prev_estim = self._get_estimations(predictions)
+    def _compute_confusion_matrix(self, predictions, y_values):
+        self.CM = confusion_matrix(y_values, predictions, labels=self.classes_).T
+        self.CM = self.CM.astype(float)
+        prev_estim = self.CM.sum(axis=0)
+
         for i, _ in enumerate(self.classes_):
             if prev_estim[i] == 0:
                 self.CM[i, i] = 1
@@ -448,13 +452,16 @@ class GPAC(SoftLearnerQMixin, MatrixAdjustment):
     def __init__(self, learner=None):
         super().__init__(learner=learner, solver='linear')
     
-    def _compute_confusion_matrix(self, posteriors):
-        prev_estim = self._get_estimations(posteriors)
-        for i, _ in enumerate(self.classes_):
-            if prev_estim[i] == 0:
-                self.CM[i, i] = 1
-            else:
-                self.CM[:, i] /= prev_estim[i]
+    def _compute_confusion_matrix(self, posteriors, y_values):
+        n_classes = len(self.classes_)
+        confusion = np.eye(n_classes)
+
+        for i, class_label in enumerate(self.classes_):
+            indices = (y_values == class_label)
+            if np.any(indices):
+                confusion[i] = posteriors[indices].mean(axis=0)
+        
+        self.CM = confusion.T
         return self.CM
 
 

{mlquantify-0.1.14.dist-info → mlquantify-0.1.16.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mlquantify
-Version: 0.1.14
+Version: 0.1.16
 Summary: Quantification Library
 Home-page: https://github.com/luizfernandolj/QuantifyML/tree/master
 Maintainer: Luiz Fernando Luth Junior

{mlquantify-0.1.14.dist-info → mlquantify-0.1.16.dist-info}/RECORD

@@ -5,7 +5,7 @@ mlquantify/calibration.py,sha256=chG3GNX2BBDTWIuSVfZUJ_YF_ZVBSoel2d_AN0OChS0,6
 mlquantify/confidence.py,sha256=QkEWr6s-Su3Nbinia_TRQbBeTM6ymDPe7Bv204XBKKA,10799
 mlquantify/multiclass.py,sha256=wFbbXKqGsFVSsI9zC0EHGYyyx1JRxFpzMi_q8l80TUM,11770
 mlquantify/adjust_counting/__init__.py,sha256=AWio99zeaUULQq9vKggkFhnq-tqgXxasQt167NdcNVY,307
-mlquantify/adjust_counting/_adjustment.py,sha256=aQ92-wfRF1TT_d3kQecFleSWE8CdGBJiZx3bbePuXM0,23284
+mlquantify/adjust_counting/_adjustment.py,sha256=412kFnx3-noaA9u9AatuGIvJbKze-PLPFfBFMBVmQVA,23635
 mlquantify/adjust_counting/_base.py,sha256=MjBsNG7wE0Z_KToXX8WbthhVvz-yc0-d2zIqPo1CB9g,9429
 mlquantify/adjust_counting/_counting.py,sha256=6PKea54xvsga8spNEbsngKNQPyGUXzOkCRyXQR8rTdo,5699
 mlquantify/adjust_counting/_utils.py,sha256=DEPNzvcr0KszCnfUJaRzBilwWzuNVMSdy5eV7aQ_JPE,2907
@@ -47,7 +47,7 @@ mlquantify/utils/_sampling.py,sha256=QQxE2WKLdiCFUfPF6fKgzyrsOUIWYf74w_w8fbYVc2c
 mlquantify/utils/_tags.py,sha256=Rz78TLpxgVxBKS0mKTlC9Qo_kn6HaEwVKNXh8pxFT7M,1095
 mlquantify/utils/_validation.py,sha256=TGGnfv7F5rnQmVeSqGMuS9AP76O974b1TPishKCCWls,16800
 mlquantify/utils/prevalence.py,sha256=FXLCJViQb2yDbyTXeGZt8WsPPnSZINhorQYZTKXOn14,1772
-mlquantify-0.1.14.dist-info/METADATA,sha256=-uPQqaXhgbXEq9M4EWYGKK9t3RRGNpbZY436DCL3bog,4701
-mlquantify-0.1.14.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-mlquantify-0.1.14.dist-info/top_level.txt,sha256=tGEkYkbbFElwULvqENjam3u1uXtyC1J9dRmibsq8_n0,11
-mlquantify-0.1.14.dist-info/RECORD,,
+mlquantify-0.1.16.dist-info/METADATA,sha256=RZWNq8k48KnN4PYd-6Iw5mv-qS6bAVL2tc6zg9cMqN0,4701
+mlquantify-0.1.16.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+mlquantify-0.1.16.dist-info/top_level.txt,sha256=tGEkYkbbFElwULvqENjam3u1uXtyC1J9dRmibsq8_n0,11
+mlquantify-0.1.16.dist-info/RECORD,,