ajout de knn pour la monoclassification et calcul du temp

code/knn.py

@@ -25,6 +25,18 @@ class Knn:
             value[i] = value[i] / ttPart
         return value
 
+    def getLabelOfPoint(self, coord,nbNearest):
+        points, dists = self.getNNearest(coord, nbNearest)
+        label = {}
+        for idx in range(len(points)):
+            tvalue = points[idx].value[0]
+            if tvalue in label:
+                label[tvalue] += 1/(dists[idx]+1)
+            else:
+                label[tvalue] = 1/(dists[idx]+1)
+        value = max(label, key=label.get)
+        return value
+
     def getNNearest(self, coord, nbNearest):
         dist = np.copy(self.space)
         dist = np.frompyfunc(lambda x: x.getDistFromCoord(coord), 1, 1)(dist)

demo4.py

@@ -2,11 +2,11 @@ from nnnar import *
 from knn import *
 import numpy as np
 import pandas as pd
-
+from time import time
 
 def runOneTest(trainTestRatio):
     nnnar = Nnnar(4, 0, 100.001, 10)# le 100.001 est pour être sûr d'être au dessus de la valeur max (100.0)
-
+    knn = Knn()
     df = pd.read_csv('./data/Iris.csv')
     df = df.iloc[:, 1:]
     # Normalisation des données
@@ -45,9 +45,14 @@ def runOneTest(trainTestRatio):
 
     return 100 - nbError / len(coord) * 100
 
+
+t1 = time()
+
 nbRepetition = 100
 accuracy = 0
 for i in range(nbRepetition):
     accuracy += runOneTest(0.8)
 
-print("accuracy moyenne: ",str(accuracy/nbRepetition))
+t2 = time()
+print("accuracy moyenne: ",str(accuracy/nbRepetition))
+print("delta temps: ",str(t2-t1))