更新 function.py

function.py

@@ -46,24 +46,26 @@ from collections import Counter
 def classify0(inX, dataSet, labels, k=1):
     """
     KNN 分类算法。
-    
+
     Args:
     - inX (numpy array): 输入待分类的样本。
     - dataSet (numpy array): 训练数据集。
     - labels (list or numpy array): 训练集对应的标签。
     - k (int): KNN 算法参数，默认为3。
-    
+
     Returns:
     - int: 预测的分类标签。
     """
-    # 计算距离
+    # 计算马氏距离
+    cov_matrix = np.cov(dataSet, rowvar=False)
+    inv_cov_matrix = np.linalg.inv(cov_matrix)
+    mean_vector = np.mean(dataSet, axis=0)
+
     diffMat = inX - dataSet
-    sqDiffMat = diffMat ** 2
-    sqDistances = sqDiffMat.sum(axis=1)
-    distances = sqDistances ** 0.5
+    mahalanobis_distances = np.sqrt(np.sum((diffMat @ inv_cov_matrix) * diffMat, axis=1))
     
     # 排序并选取前k个最小距离
-    sortedDistIndices = distances.argsort()
+    sortedDistIndices = mahalanobis_distances.argsort()
     classCount = {}
 
     for i in range(k):
@@ -83,40 +85,13 @@ def load_data(file_path):
     labels = data[:, -1].astype(int)
     return features, labels
 
-def normalize_min_max(data):
-    """
-    对数据进行最小-最大缩放归一化。
-    
-    Args:
-    - data (numpy array): 待归一化的数据。
-    
-    Returns:
-    - numpy array: 归一化后的数据。
-    """
-    min_vals = data.min(0)
-    max_vals = data.max(0)
-    ranges = max_vals - min_vals
-    # 避免除以零的情况
-    ranges[ranges == 0.0] = 1.0
-    m = data.shape[0]
-    norm_data = data - np.tile(min_vals, (m, 1))
-    norm_data = norm_data / np.tile(ranges, (m, 1))
-    return norm_data, min_vals, max_vals
-
 if __name__ == "__main__":
     # 以下代码可用于测试，可自行随意修改
     train_file = "datingTrainSet2.tsv"
     
-    # 加载训练数据
     train_data, train_labels = load_data(train_file)
 
-    # 归一化训练数据，并获取最小值和最大值
-    train_data_normalized, min_vals, max_vals = normalize_min_max(train_data)
-
-    # 测试数据点也需要用同样的方式归一化
+    # 测试数据点
     test_point = np.array([65915, 3.868359, 0.128078])
-    test_point_normalized = (test_point - min_vals) / (max_vals - min_vals)
-
-    # 使用归一化后的数据进行预测
-    predicted_label = classify0(test_point_normalized, train_data_normalized, train_labels)
-    print("预测的分类标签是:", predicted_label)
+    predicted_label = classify0(test_point, train_data, train_labels)
+    print("预测的分类标签是:", predicted_label)