更新 function.py

function.py

@@ -28,45 +28,36 @@ name = "章程鸿"
 # 这里写自己的代码实现
 import numpy as np
 
+
 def sigmoid(z):
     return 1 / (1 + np.exp(-z))
 
-def train_logreg(X, y, learning_rate=0.01, iterations=1000, batch_size=32, init_weights=None):
+
+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。
-    - batch_size (int): 小批量大小，默认为 32。
-    - init_weights (numpy array): 权重初始值，默认为 None（即全零初始化）。
+    - other_params (str): 其它参数，可选设置。若需设置其它参数，必须设置其默认值，否则将**无法通过**自动评测，**成绩为零**。
 
     Returns:
     - numpy array: 训练得到的权重。
     """
     m, n = X.shape
-    if init_weights is None:
-        weights = np.zeros(n)
-    else:
-        weights = init_weights
+    weights = np.zeros(n)
     bias = 0
     for _ in range(iterations):
-        indices = np.random.permutation(m)
-        X_shuffled = X[indices]
-        y_shuffled = y[indices]
-        for i in range(0, m, batch_size):
-            X_batch = X_shuffled[i:i+batch_size]
-            y_batch = y_shuffled[i:i+batch_size]
-            z = np.dot(X_batch, weights) + bias
-            h = sigmoid(z)
-            dw = (1/batch_size) * np.dot(X_batch.T, (h - y_batch))
-            db = (1/batch_size) * np.sum(h - y_batch)
-            weights -= learning_rate * dw
-            bias -= learning_rate * db
+        z = np.dot(X, weights) + bias
+        h = sigmoid(z)
+        dw = (1 / m) * np.dot(X.T, (h - y))
+        db = (1 / m) * np.sum(h - y)
+        weights -= learning_rate * dw
+        bias -= learning_rate * db
     return weights
 
+
 def predict_logreg(inX, weights):
     """
     对率回归预测函数。
@@ -82,6 +73,7 @@ def predict_logreg(inX, weights):
     h = sigmoid(z)
     return 1 if h >= 0.5 else 0
 
+
 def load_data(filepath):
     """
     从文件中加载数据，若不自行正确实现，则无法通过评测，**不记录成绩**。
@@ -95,31 +87,55 @@ def load_data(filepath):
     """
     data = []
     with open(filepath, 'r') as file:
-        for line in file:
-            # 清理无效数据
-            cleaned_line = line.strip().replace('?', '0')
-            values = [float(x) for x in cleaned_line.split()]
-            # 处理目标变量
-            if values[22] in [2, 3]:
-                values[22] = 0
-            new_values = values[:2] + values[3:23]
+        lines = file.readlines()
+        # 处理每一行数据
+        for line in lines:
+            values = line.strip().split()
+            new_values = []
+            for i in range(len(values)):
+                if values[i] == '?':
+                    if i == 0 or i == 1:
+                        new_values.append(0)
+                    else:
+                        new_values.append(0)  # 先暂时用0填充，后续修改为均值填充
+                else:
+                    new_values.append(float(values[i]))
             data.append(new_values)
     data = np.array(data)
     X = data[:, :-1]
     y = data[:, -1]
     return X, y
 
+
+def fill_with_mean(X):
+    """
+    用均值填充特征矩阵中的缺失值
+
+    Args:
+    - X (numpy array): 特征数据
+
+    Returns:
+    - numpy array: 填充后的特征数据
+    """
+    for i in range(X.shape[1]):
+        col = X[:, i]
+        mean_val = np.mean(col[col!= 0])
+        col[col == 0] = mean_val
+    return X
+
+
 if __name__ == "__main__":
     # 加载训练集
     X_train, y_train = load_data("./data/horse-colic.data")
+    X_train = fill_with_mean(X_train)
 
     # 加载测试集
     X_test, y_test = load_data("./data/horse-colic.test")
+    X_test = fill_with_mean(X_test)
 
     # 训练对率回归模型
-    init_weights = np.random.randn(X_train.shape[1])  # 随机初始化权重
-    weights = train_logreg(X_train, y_train, learning_rate=0.01, iterations=1000, batch_size=32, init_weights=init_weights)
+    weights = train_logreg(X_train, y_train, learning_rate=0.01, iterations=1000)
 
     # 测试单个数据
     y_pred = predict_logreg(X_test[11], weights)
-    print(f"预测结果: {y_pred}, 实际结果: {y_test[11]}")
+    print(f"预测结果: {y_pred}, 实际结果: {y_test[11]}")