Merge pull request #348 from MartianCoder-git/patch-1

KNN_algorithm

Author: seeditsolution

KNN_algorithm

@@ -0,0 +1,65 @@
+from sklearn import datasets
+from sklearn.model_selection import train_test_split
+from sklearn.neighbors import KNeighborsClassifier
+from sklearn.metrics import accuracy_score
+from scipy.spatial import distance
+
+def euc(a,b):
+    return distance.euclidean(a, b)
+   
+
+class ScrappyKNN():
+    def fit(self, features_train, labels_train):
+        self.features_train = features_train
+        self.labels_train = labels_train
+       
+    def predict(self, features_test):
+        predictions = []
+        for item in features_test:
+            label = self.closest(item)
+            predictions.append(label)
+           
+        return predictions
+   
+    def closest(self, item):
+       best_dist = euc(item, self.features_train[0])
+       best_index = 0
+       for i in range(1,len(self.features_train)):
+           dist = euc(item, self.features_train[i])
+           if dist < best_dist:
+               best_dist = dist
+               best_index = i
+       return self.labels_train[best_index]
+
+iris = datasets.load_iris()
+
+print(iris)
+
+features = iris.data
+labels = iris.target
+
+print(features)
+print(labels)
+
+features_train, features_test, labels_train, labels_test  = train_test_split(features, labels, test_size=.5)
+#print(len(features))
+#print(len(features_train))
+
+my_classifier = ScrappyKNN()
+#my_classifier = KNeighborsClassifier()
+my_classifier.fit(features_train, labels_train)
+
+prediction = my_classifier.predict(features_test)
+
+print(prediction)
+print(accuracy_score(labels_test, prediction))
+
+iris1 = [[7.1, 2.9, 5.3, 2.4]] #virginica
+iris_prediction = my_classifier.predict(iris1)
+
+if iris_prediction == 0:
+    print("Setosa")
+if iris_prediction == 1:
+    print("Versicolor")
+if iris_prediction == 2:
+    print("Virginica")