From 4470d0c92542fdda7aba42b0b925d6552db7e50c Mon Sep 17 00:00:00 2001
From: ywq157 <271203576@qq.com>
Date: Thu, 21 Nov 2024 15:29:57 +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 | 30 +++++++++++++++++++++++-------
 1 file changed, 23 insertions(+), 7 deletions(-)

diff --git a/function.py b/function.py
index 06aaed6..af11423 100644
--- a/function.py
+++ b/function.py
@@ -31,6 +31,7 @@ name = "杨文强"
 def load_data(filepath):
     """
     从文件中加载数据，并返回特征和标签。
+    如果数据中包含 '?' 字符，将其处理为 NaN。
     
     Args:
     - filepath (str): 数据文件路径。
@@ -39,19 +40,34 @@ def load_data(filepath):
     - np.array: 特征数据。
     - np.array: 标签数据。
     """
-    data = np.loadtxt(filepath, delimiter=" ", dtype=str)
+    # 读取数据，指定将 '?' 转换为 np.nan
+    data = np.genfromtxt(filepath, delimiter=" ", dtype=str, missing_values='?', filling_values=np.nan)
+
     X = []
     y = []
 
     for row in data:
-        features = row[:-1].astype(float)  # 特征数据
-        label = int(row[-1])  # 标签数据 (0 或 1)
-        
-        # 转换标签为 +1 或 -1
-        y.append(1 if label == 1 else -1)
+        features = row[:-1]  # 特征数据
+        label = row[-1]      # 标签数据
+
+        # 将特征中的 '?' 替换为 NaN，后续可以选择删除这些行或填充NaN值
+        features = [float(x) if x != '?' else np.nan for x in features]
+
+        # 标签的处理：确保标签是 1 或 -1
+        label = int(label)  # 标签需要转换为整数 1 或 0
+        y.append(1 if label == 1 else -1)  # 将标签转换为 +1 或 -1
         X.append(features)
     
-    return np.array(X), np.array(y)
+    X = np.array(X)
+    y = np.array(y)
+
+    # 处理缺失值：可以选择删除含有NaN的样本
+    # 这里是删除含有NaN的行
+    valid_indices = ~np.isnan(X).any(axis=1)
+    X = X[valid_indices]
+    y = y[valid_indices]
+
+    return X, y
 
 def train_svm(X, y, learning_rate=0.01, iterations=1000, regularization_param=0.01):
     """