From eff8dfd6d928818d1a1950be133a32a9aedaf016 Mon Sep 17 00:00:00 2001
From: 202210423104 <1961532386@qq.com>
Date: Wed, 16 Oct 2024 22:10:25 +0800
Subject: [PATCH] =?UTF-8?q?=E4=B8=8A=E4=BC=A0=E6=96=87=E4=BB=B6=E8=87=B3?=
 =?UTF-8?q?=20/?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 function.py | 42 +++++++++++++++++++++++++++---------------
 1 file changed, 27 insertions(+), 15 deletions(-)

diff --git a/function.py b/function.py
index 630511a..946a307 100644
--- a/function.py
+++ b/function.py
@@ -5,7 +5,8 @@ from collections import Counter
 id = "202210423104"
 name = "张钦诚"
 
-def classify0(inX, dataSet, labels, k=5):
+
+def classify0(inX, dataSet, labels, k=1):
     """
     加权 kNN 分类算法。
 
@@ -13,7 +14,7 @@ def classify0(inX, dataSet, labels, k=5):
     - inX (list or 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，默认值为5。
+    - k (int): kNN 算法参数，默认值为1。
 
     Returns:
     - int: 预测的分类标签。
@@ -24,13 +25,23 @@ def classify0(inX, dataSet, labels, k=5):
     # 获取距离最小的k个样本的索引
     nearest_indices = np.argpartition(distances, k)[:k]
     nearest_labels = labels[nearest_indices]
-    nearest_distances = distances[nearest_indices]
-    # 计算权重，距离越近权重越大
-    weights = 1 / (nearest_distances + 1e-5)  # 防止除以零
-    # 使用 np.bincount 进行加权统计
-    label_weights = np.bincount(nearest_labels, weights=weights)
-    # 返回权重最大的标签
-    return np.argmax(label_weights)
+
+    if k == 1:
+        # 如果k=1，直接返回最近邻的标签
+        return nearest_labels[0]
+    else:
+        # 加权 kNN
+        nearest_distances = distances[nearest_indices]
+        weights = 1 / (nearest_distances + 1e-5)  # 防止除以零
+        label_weights = {}
+        for label, weight in zip(nearest_labels, weights):
+            if label in label_weights:
+                label_weights[label] += weight
+            else:
+                label_weights[label] = weight
+        # 返回权重最大的标签
+        return max(label_weights.items(), key=lambda x: x[1])[0]
+
 
 def load_data(file_path):
     """
@@ -47,6 +58,7 @@ def load_data(file_path):
     labels = data[:, -1].astype(int)
     return features, labels
 
+
 def standardize(train_data, test_data):
     """
     对训练数据和测试数据进行标准化处理。
@@ -56,15 +68,15 @@ def standardize(train_data, test_data):
     - test_data (numpy array): 测试数据集。
 
     Returns:
-    - tuple: (标准化后的训练数据, 标准化后的测试数据)
+    - tuple: (标准化后的训练数据, 标准化后的测试数据, 均值, 标准差)
     """
     mean = np.mean(train_data, axis=0)
     std = np.std(train_data, axis=0)
-    # 防止标准差为零
-    std[std == 0] = 1
+    std[std == 0] = 1  # 防止标准差为零
     train_standardized = (train_data - mean) / std
     test_standardized = (test_data - mean) / std
-    return train_standardized, test_standardized
+    return train_standardized, test_standardized, mean, std
+
 
 if __name__ == "__main__":
     train_file = "datingTrainSet2.tsv"
@@ -76,12 +88,12 @@ if __name__ == "__main__":
     test_data, test_labels = load_data(test_file)
 
     # 数据标准化
-    train_data, test_data = standardize(train_data, test_data)
+    train_data, test_data, mean, std = standardize(train_data, test_data)
 
     # 测试样本
     test_sample = [65915, 3.868359, 0.128078]
     # 标准化测试样本
-    test_sample = (np.array(test_sample) - np.mean(train_data, axis=0)) / np.std(train_data, axis=0)
+    test_sample = (np.array(test_sample) - mean) / std
 
     # 预测标签
     predicted_label = classify0(test_sample, train_data, train_labels)