From 8aa000191460f8f3f18de52727a4ff37314780c9 Mon Sep 17 00:00:00 2001
From: Frae19 <515745323@qq.com>
Date: Thu, 17 Oct 2024 16:58:13 +0800
Subject: [PATCH] =?UTF-8?q?=E6=9B=B4=E6=96=B0=20function.py?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 function.py | 13 +++++--------
 1 file changed, 5 insertions(+), 8 deletions(-)

diff --git a/function.py b/function.py
index 3a48f96..8877e42 100644
--- a/function.py
+++ b/function.py
@@ -57,24 +57,21 @@ def classify0(inX, dataSet, labels, k=3):
     # 计算欧式距离  
     dataSetSize = dataSet.shape[0]  
     diffMat = np.tile(inX, (dataSetSize, 1)) - dataSet  
-    sqDiffMat = diffMat**2  
-    sqDistances = sqDiffMat.sum(axis=1)  
-    distances = np.sqrt(sqDistances)  
+    distances = np.sqrt(np.sum(diffMat ** 2, axis=1))  
     
     # 按距离从小到大排序  
     sortedDistIndices = distances.argsort()  
-    
-    # 选择距离最小的k个点  
+
     classCount = {}  
     for i in range(k):  
         voteIlabel = labels[sortedDistIndices[i]]  
-        classCount[voteIlabel] = classCount.get(voteIlabel, 0) + 1  
+        weight = 1 / (distances[sortedDistIndices[i]] + 1e-8)  # 防止除以零，加入微小值  
+        classCount[voteIlabel] = classCount.get(voteIlabel, 0) + weight  
     
     # 按类别票数从大到小排序  
     sortedClassCount = sorted(classCount.items(), key=lambda x: x[1], reverse=True)  
     
-    # 返回出现频率最高的类别  
-    return sortedClassCount[0][0]  
+    return sortedClassCount[0][0]
 
 
 def load_data(file_path):