更新 function.py

function.py

@@ -1,5 +1,8 @@
-# function.py
 import numpy as np
+from sklearn.impute import SimpleImputer
+from sklearn.preprocessing import StandardScaler
+import csv
+
 """
 实验说明:
 - 自行编写实现数据加载与预测病马死亡的对率回归算法。
@@ -18,92 +21,88 @@ import numpy as np
 
 
 """
-202210423219李燕如
+请在这里填写学号和姓名，不填写即成绩无效。
 """
 id = "202210423219"
 name = "李燕如"
 
 
 # 这里写自己的代码实现
-def load_data(filepath):
+def sigmoid(x):
-    """
+    return 1 / (1 + np.exp(-x))
-    从文件中加载数据。
-
-    Args:
-    - filepath (str): 数据文件路径。
-
-    Returns:
-    - np.array: 特征数据。
-    - np.array: 标签数据。
-    """
-    data = np.genfromtxt(filepath, delimiter=',', skip_header=1, filling_values=np.nan)
-    
-    # 分离特征和标签
-    X = data[:, :-1]
-    y = data[:, -1].flatten()  # 确保y是一维数组
-
-    # 删除包含NaN的行
-    mask = ~np.isnan(X).any(axis=1) & ~np.isnan(y)
-    X = X[mask]
-    y = y[mask]
-    
-    return X, y
-
-def sigmoid(z):
-    """
-    Sigmoid函数。
-
-    Args:
-    - z (numpy array): 输入值。
-
-    Returns:
-    - numpy array: Sigmoid函数的输出值。
-    """
-    return 1 / (1 + np.exp(-z))
-
-def train_logreg(X, y, learning_rate=0.01, iterations=1000):
-    """
-    对率回归训练函数。
-
-    Args:
-    - X (numpy array): 特征数据。
-    - y (numpy array): 标签数据。
-    - learning_rate (float): 学习率，默认为0.01。
-    - iterations (int): 迭代次数，默认为1000。
-
-    Returns:
-    - numpy array: 训练得到的权重。
-    """
-    # 添加截距项
-    X = np.c_[np.ones(X.shape[0]), X]
-    # 初始化权重
-    weights = np.zeros(X.shape[1])
-
-    # 检查y是否为空
-    if len(y) == 0:
-        raise ValueError("标签数组y为空")
 
+def train_logreg(X, y, learning_rate=0.01, iterations=1000, regularization=0.01):
+    m, n = X.shape
+    weights = np.zeros(n)
     for _ in range(iterations):
-        # 计算预测值
+        linear_model = np.dot(X, weights)
-        predictions = sigmoid(np.dot(X, weights))
+        y_predicted = sigmoid(linear_model)
-        # 计算梯度
+        gradient = np.dot(X.T, (y_predicted - y)) / m + regularization * weights
-        gradient = np.dot(X.T, (predictions - y)) / max(len(y), 1)  # 防止除以零
-        # 更新权重
         weights -= learning_rate * gradient
-
     return weights
 
 def predict_logreg(inX, weights):
-    """
+    linear_output = np.dot(inX, weights)
-    对率回归预测函数。
+    probability = sigmoid(linear_output)
+    return 1 if probability > 0.5 else 0
 
-    Args:
+def preprocess_data(X, y):
-    - inX (numpy array): 输入待分类的样本。
+    # 合并目标变量的类别
-    - weights (numpy array): 训练得到的权重。
+    y[y == 2] = 0
+    y[y == 3] = 0
 
-    Returns:
+    # 删除特征3和特征24~28
-    - int: 预测的分类标签。
+    X = np.delete(X, [3], axis=1)
-    """
+    X = X[:, :23]
-    inX = np.c_[1, inX]  # 添加截距项
+
-    prob = sigmoid(np.dot(inX, weights))
+    return X, y
-    return 1 if prob > 0.5 else 0
+
+def load_data(filepath):
+    # 使用csv模块读取数据
+    data = []
+    with open(filepath, 'r') as file:
+        reader = csv.reader(file, delimiter=' ')
+        for row in reader:
+            if row and row[0] != '?':  # 忽略以'?'开头的行
+                row = [float(x) if x != '?' else np.nan for x in row]
+                data.append(row)
+    
+    # 将'?'替换为np.nan
+    data = np.array(data, dtype=float)
+
+    # 使用SimpleImputer填充缺失值
+    imputer = SimpleImputer(missing_values=np.nan, strategy='constant', fill_value=0)
+    data = imputer.fit_transform(data)
+
+    # 将所有列转换为数值类型
+    X = data[:, :-1]
+    y = data[:, -1].astype(int)
+
+    # 预处理数据
+    X, y = preprocess_data(X, y)
+
+    return X, y
+
+def main():
+    # 更新文件路径以匹配实际位置
+    X_train, y_train = load_data("C:/Users/HP/Desktop/机器学习/Experiment-LogReg/data/horse-colic.data")
+    X_test, y_test = load_data("C:/Users/HP/Desktop/机器学习/Experiment-LogReg/data/horse-colic.test")
+
+    if X_train is not None and X_test is not None:
+        # 特征缩放
+        scaler = StandardScaler()
+        X_train = scaler.fit_transform(X_train)
+        X_test = scaler.transform(X_test)
+
+        # 训练模型
+        weights = train_logreg(X_train, y_train, learning_rate=0.01, iterations=1000, regularization=0.1)
+
+        # 预测测试集
+        y_pred = [predict_logreg(x, weights) for x in X_test]
+
+        # 计算准确率
+        accuracy = np.mean(y_pred == y_test)
+        print(f"准确率: {accuracy}")
+        
+if __name__ == "__main__":
+    main()