更新 function.py

function.py

@@ -37,15 +37,9 @@ name = "吴俊煯"
 """
 
 
-def euclidean_distance(inX, dataSetRow):  
-    """  
-    计算欧氏距离  
-    """  
-    return np.sqrt(np.sum((inX - dataSetRow) ** 2))
-
 
 # 这里写自己的代码实现
-def classify0(inX, dataSet, labels,k=3):
+def classify0(inX, dataSet, labels,k=4):
     """
     KNN 分类算法。
 
@@ -57,21 +51,26 @@ def classify0(inX, dataSet, labels,k=3):
     Returns:
     - int: 预测的分类标签。
     """
-    # 计算每个训练样本与输入样本的欧氏距离  
+        # 确保 k 不大于数据集的大小  
-    distances = [euclidean_distance(inX, dataSet[i]) for i in range(len(dataSet))]  
+    k = min(k, dataSet.shape[0])  
       
-    # 排序并获取索引  
+    # 计算测试样本与每个训练样本之间的距离  
-    sorted_distances_indices = np.argsort(distances)  
+    diffMat = np.tile(inX, (dataSet.shape[0], 1)) - dataSet  
+    sqDiffMat = diffMat ** 2  
+    distances = np.sqrt(sqDiffMat.sum(axis=1))  
       
-    # 选择前 k 个距离最小的样本  
+    # 对距离进行排序，并获取索引  
-    k_indices = sorted_distances_indices[:k]  
+    sortedDistIndicies = distances.argsort()  
       
-    # 提取这些样本的标签  
+    # 选择距离最小的k个点  
-    k_nearest_labels = [labels[i] for i in k_indices]  
+    classCount = {}  
+    for i in range(k):  
+        voteIlabel = labels[sortedDistIndicies[i]]  
+        classCount[voteIlabel] = classCount.get(voteIlabel, 0) + 1  
       
-    # 对这些标签进行投票，返回出现次数最多的标签  
+    # 多数投票  
-    from collections import Counter  
+    sortedClassCount = sorted(classCount.items(), key=lambda item: item[1], reverse=True)  
-    return Counter(k_nearest_labels).most_common(1)[0][0]
+    return sortedClassCount[0][0]
 
 
 def load_data(file_path):