Enhance code

debug.cpp

@@ -0,0 +1,14 @@
+std::cout << "+++++++++++++++++++++++" << std::endl;
+for (size_t i = 0; i < y.size(); i++)
+{
+    printf("(%3.1f, %d)\n", X[indices.at(i)], y[indices.at(i)]);
+}
+std::cout << "+++++++++++++++++++++++" << std::endl;
+
+std::cout << "Information Gain:" << std::endl;
+auto nc = Metrics::numClasses(y, indices, 0, indices.size());
+for (auto cutPoint = cutIdx.begin(); cutPoint != cutIdx.end(); ++cutPoint)
+{
+    std::cout << *cutPoint << " -> " << Metrics::informationGain(y, indices, 0, indices.size(), *cutPoint, nc) << std::endl;
+    //  << Metrics::informationGain(y, 0, y.size(), *cutPoint, Metrics::numClasses(y, 0, y.size())) << std::endl;
+}

fimdlp/CPPFImdlp.cpp

@@ -53,20 +53,6 @@ namespace CPPFImdlp
             yPrev = y.at(*index);
             idxPrev = *index;
         }
-        std::cout << "Information Gain:" << std::endl;
-        auto nc = Metrics::numClasses(y, indices, 0, indices.size());
-        for (auto cutPoint = cutIdx.begin(); cutPoint != cutIdx.end(); ++cutPoint)
-        {
-            std::cout << *cutPoint << " -> " << Metrics::informationGain(y, indices, 0, indices.size(), *cutPoint, nc) << std::endl;
-            //  << Metrics::informationGain(y, 0, y.size(), *cutPoint, Metrics::numClasses(y, 0, y.size())) << std::endl;
-        }
-        std::cout << "+++++++++++++++++++++++" << std::endl;
-        for (size_t i = 0; i < y.size(); i++)
-        {
-            printf("(%3.1f, %d)\n", X[indices.at(i)], y[indices.at(i)]);
-        }
-        std::cout << "+++++++++++++++++++++++" << std::endl;
-
         return cutPts;
     }
     // Argsort from https://stackoverflow.com/questions/1577475/c-sorting-and-keeping-track-of-indexes

fimdlp/mdlp.py

@@ -62,7 +62,6 @@ class FImdlp(TransformerMixin, BaseEstimator):
         self.X_ = X
         self.y_ = y
         self.discretizer_ = CFImdlp()
-
         return self
 
     def transform(self, X):
@@ -92,7 +91,7 @@ class FImdlp(TransformerMixin, BaseEstimator):
             raise ValueError(
                 "Shape of input is different from what was seen in `fit`"
             )
-        print("Cut points for each feature in Iris dataset:")
+        print("Calculating cut points in python for first feature")
         yz = self.y_.copy()
         xz = X[:, 0].copy()
         xz = xz[np.argsort(X[:, 0])]
@@ -102,14 +101,12 @@ class FImdlp(TransformerMixin, BaseEstimator):
             if yz[i] != yz[i - 1] and xz[i - 1] < xz[i]:
                 print(f"Cut point: ({xz[i-1]}, {xz[i]}) ({yz[i-1]}, {yz[i]})")
                 cuts.append((xz[i] + xz[i - 1]) / 2)
-        for i in range(0, 1):  # self.n_features_):
-            datax = np.sort(X[:, i])
-            Xcutpoints = self.discretizer_.cut_points(datax, self.y_)
+        print("Cuts calculados en python: ", cuts)
+        print("Cuts calculados en C++")
+        print("Cut points for each feature in Iris dataset:")
+        for i in range(0, self.n_features_):
+            datax = self.X_[np.argsort(self.X_[:, i]), i]
+            y_ = self.y_[np.argsort(self.X_[:, i])]
+            Xcutpoints = self.discretizer_.cut_points(datax, y_)
             print(f"{self.features_[i]:20s}: {Xcutpoints}")
-        print("Solución cut_points: ", cuts)
-        print(xz)
-        print("***********")
-        for i in range(0, len(yz)):
-            print(f"({xz[i]}, {yz[i]})")
-        print("***********")
         return X

sample.py

@@ -7,9 +7,4 @@ X = data.data
 y = data.target
 features = data.feature_names
 test = FImdlp()
-Xcutpoints = test.fit(X, y, features=features).transform(X)
-clf = CFImdlp(debug=True)
-print("Cut points for feature 0 in Iris dataset:")
-print(clf.cut_points(X[:, 0], y))
-print("Xcut")
-print(Xcutpoints)
+test.fit(X, y, features=features).transform(X)