更新 function.py

function.py

@@ -42,6 +42,8 @@ name = "黎世浩"
 
 # 这里写自己的代码实现
 
+import numpy as np
+
 def classify0(inX, dataSet, labels, k=3):
     """
     KNN 分类算法，使用马氏距离。
@@ -55,27 +57,25 @@ def classify0(inX, dataSet, labels, k=3):
     Returns:
     - int: 预测的分类标签。
     """
-    # 计算协方差矩阵的逆和均值
+    # 计算训练数据集的协方差矩阵和均值
-    inv_cov = np.linalg.inv(np.cov(dataSet.T))  # 计算协方差矩阵的逆
+    covMat = np.cov(dataSet, rowvar=False)
-    meanVals = np.mean(dataSet, axis=0)  # 计算均值
+    meanVec = np.mean(dataSet, axis=0)
     
-    # 计算输入样本与训练集中每个样本的马氏距离
+    # 计算待分类样本与训练集中每个样本的马氏距离
-    diffMat = dataSet - meanVals
+    invCovMat = np.linalg.inv(covMat)
-    mahalDist = np.array([np.sqrt((inX - meanVals).dot(inv_cov).dot((inX - meanVals).T)) * np.sqrt(np.sum((diff - meanVals).dot(inv_cov).dot(diff - meanVals.T)) / (diffMat.shape[0] - 1)) for diff in diffMat])
+    diffVecs = dataSet - meanVec
+    dist = np.sqrt(np.sum(inX - meanVec, axis=1).dot(invCovMat).dot((inX - meanVec).T))
     
-    # 确保 k 不大于训练集中的样本数量
+    # 获取距离最近的 k 个样本的索引
-    k = min(k, len(mahalDist))
+    kNearestIndices = np.argsort(dist)[:k]
     
-    # 对距离进行排序，选择距离最小的 k 个样本
+    # 统计这些邻居的标签
-    sortedDistIndices = np.argsort(mahalDist)
     classCount = {}
+    for i in kNearestIndices:
+        voteLabel = labels[i]
+        classCount[voteLabel] = classCount.get(voteLabel, 0) + 1
     
-    # 遍历距离最小的 k 个样本，统计每个类别的出现次数
+    # 返回出现次数最多的标签
-    for i in range(k):
-        voteIlabel = labels[sortedDistIndices[i]]
-        classCount[voteIlabel] = classCount.get(voteIlabel, 0) + 1
-
-    # 选择出现次数最多的类别作为预测标签
     sortedClassCount = sorted(classCount.items(), key=lambda x: x[1], reverse=True)
     return sortedClassCount[0][0]