更新 function.py

function.py

@@ -1,4 +1,9 @@
-import numpy as np
+import numpy as npimport numpy as np
+import pandas as pd
+from sklearn.impute import SimpleImputer
+from sklearn.preprocessing import StandardScaler
+
+
 """
 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。
@@ -24,73 +29,75 @@ name = "陈瑞琪祥"
 
 
 
-def sigmoid(z):
+# Sigmoid 函数
-    return 1 / (1 + np.exp(-z))
+def sigmoid(x):
+    return 1 / (1 + np.exp(-x))
 
+# 逻辑回归训练函数
 def train_logreg(X, y, learning_rate=0.01, iterations=1000, regularization=0.01):
     m, n = X.shape
     weights = np.zeros(n)
-    for i in range(iterations):
+    for _ in range(iterations):
         linear_model = np.dot(X, weights)
         y_predicted = sigmoid(linear_model)
-        gradient = np.dot(X.T, (y_predicted - y)) / m
+        gradient = np.dot(X.T, (y_predicted - y)) / m + regularization * weights
-        # 添加L2正则化
-        gradient += regularization * np.r_[[0], weights[1:]]  # 不对偏置项进行正则化
         weights -= learning_rate * gradient
     return weights
 
+# 逻辑回归预测函数
 def predict_logreg(inX, weights):
     linear_output = np.dot(inX, weights)
     probability = sigmoid(linear_output)
     return 1 if probability > 0.5 else 0
 
-def load_data(filepath):
+# 数据预处理函数
-    """
+def preprocess_data(X, y):
-    从文件中加载数据。
+    # 合并目标变量的类别
+    y[y == 2] = 0
+    y[y == 3] = 0
 
-    Args:
+    # 删除特征3和特征24~28
-    - filepath (str): 数据文件路径。
+    X = np.delete(X, [3], axis=1)
-
+    X = X[:, :23]
-    Returns:
-    - X (np.array): 特征数据。
-    - y (np.array): 标签数据。
-    """
-    # 使用numpy的genfromtxt函数读取数据，以空格为分隔符
-    data = np.genfromtxt(filepath, delimiter=' ', invalid_raise=False, filling_values=np.nan)
-    
-    # 检查数据是否为一维数组
-    if data.ndim == 1:
-        print("Data file may have an incorrect delimiter or only contains one column.")
-        return None, None
-
-    # 丢弃完全为NaN的行
-    data = data[~np.isnan(data).any(axis=1)]
-
-    # 分离特征和标签
-    X = data[:, :-1]
-    y = data[:, -1]
 
     return X, y
 
-def feature_scaling(X):
+# 加载数据函数
-    # 特征归一化
+def load_data(filepath):
-    X_min = X.min(axis=0)
+    try:
-    X_max = X.max(axis=0)
+        data = pd.read_csv(filepath, header=None, sep=r'\s+', on_bad_lines='skip')
-    X_scaled = (X - X_min) / (X_max - X_min)
+    except pd.errors.ParserError as e:
-    return X_scaled
+        print(f"Error parsing file: {e}")
+        return None, None
+
+    data.replace('?', np.nan, inplace=True)
+    imputer = SimpleImputer(missing_values=np.nan, strategy='constant', fill_value=0)
+    data.iloc[:, :-1] = imputer.fit_transform(data.iloc[:, :-1])
+
+    for col in data.columns[:-1]:
+        data[col] = pd.to_numeric(data[col], errors='coerce')
+
+    X = data.iloc[:, :-1].values
+    y = data.iloc[:, -1].values.astype(int)
+
+    X, y = preprocess_data(X, y)
+    return X, y
+
+# 主函数
+def main():
+    X_train, y_train = load_data("D:/data/horse/horse-colic.data")
+    X_test, y_test = load_data("D:/data/horse/horse-colic.test")
+
+    if X_train is not None and X_test is not None:
+        scaler = StandardScaler()
+        X_train = scaler.fit_transform(X_train)
+        X_test = scaler.transform(X_test)
+
+        weights = train_logreg(X_train, y_train, learning_rate=0.01, iterations=1000, regularization=0.1)
+        y_pred = [predict_logreg(x, weights) for x in X_test]
+
+        accuracy = np.mean(y_pred == y_test)
+        print(f"准确率: {accuracy}")
 
 if __name__ == "__main__":
-    X_train, y_train = load_data("./data/horse-colic.data")
+    main()
-    X_test, y_test = load_data("./data/horse-colic.test")
-
-    # 特征归一化
-    X_train_scaled = feature_scaling(X_train)
-    X_test_scaled = feature_scaling(X_test)
-
-    # 训练对率回归模型
-    weights = train_logreg(X_train_scaled, y_train, learning_rate=0.01, iterations=1000)
-
-    # 测试单个数据
-    y_pred = predict_logreg(X_test_scaled[11], weights)
-    print(f"预测结果: {y_pred}, 实际结果: {y_test[11]}")
-