更新 function.py

function.py

@@ -43,18 +43,21 @@ name = "赵昊龙"
 # 这里写自己的代码实现
 import numpy as np
 
-def mahalanobis_distance(inX, dataSet, cov_mat):
-  
+def mahalanobis_distance(inX, dataSet, inv_cov_mat):
+    
     inX = np.array(inX).reshape(1, -1)
     diff = inX - dataSet
-    distances = np.sqrt(diff @ cov_mat @ diff.T)
+    distances = np.sqrt(np.sum(diff @ inv_cov_mat * diff, axis=1))
     return distances
 
 def classify0(inX, dataSet, labels, k=5):
+    
     # 计算训练数据集的协方差矩阵
     cov_mat = np.cov(dataSet.T)
+    # 计算协方差矩阵的逆矩阵
+    inv_cov_mat = np.linalg.inv(cov_mat)
     # 计算待分类样本与数据集中每个样本的马氏距离
-    distances = mahalanobis_distance(inX, dataSet, cov_mat)
+    distances = mahalanobis_distance(inX, dataSet, inv_cov_mat)
     # 获取距离最小的k个样本的索引
     sortedDistIndices = distances.argsort()
     # 为分类计数器初始化
@@ -69,9 +72,10 @@ def classify0(inX, dataSet, labels, k=5):
     return sortedClassCount[0][0]
 
 def load_data(file_path):
+    
     data = np.loadtxt(file_path, delimiter="\t")
     features = data[:, :-1]
-    labels = data[:, -1].astype(int)
+    labels = data[:, -1]
     return features, labels
 
 if __name__ == "__main__":
@@ -79,3 +83,4 @@ if __name__ == "__main__":
     train_file = "datingTrainSet2.tsv"
     train_data, train_labels = load_data(train_file)
     print(classify0([65915, 3.868359, 0.128078], train_data, train_labels))
+