更新 function.py

function.py

@@ -26,10 +26,6 @@ name = "刘子豪"
 
 
 # 这里写自己的代码实现
-def sigmoid(z):
-    return 1 / (1 + np.exp(-z)) #用于将数据转化成一个0 - 1之间的概率值
-
-
 def train_logreg(X, y, other_params="你自己设置一些其它参数"):
     """
     对率回归训练函数。
@@ -44,11 +40,14 @@ def train_logreg(X, y, other_params="你自己设置一些其它参数"):
     """
     m, n = X.shape
     weights = np.zeros(n)
+
     for _ in range(iterations):
-        predictions = sigmoid(np.dot(X, weights))
+        z = np.dot(X, weights)
+        predictions = 1 / (1 + np.exp(-z))  # 计算sigmoid
         errors = y - predictions
         gradient = np.dot(X.T, errors) / m
         weights += learning_rate * gradient
+    
     return weights
 
 
@@ -63,7 +62,8 @@ def predict_logreg(inX, weights):
     Returns:
     - int: 预测的分类标签。
     """
-    probability = sigmoid(np.dot(inX, weights))
+    z = np.dot(inX, weights)
+    probability = 1 / (1 + np.exp(-z))
     return 1 if probability >= 0.5 else 0
 
 
@@ -84,17 +84,14 @@ def load_data(filepath):
     with open(filepath, 'r') as file:
         for line in file:
             fields = line.strip().split()
-            # 检查是否有缺失值，并排除不符合行长度要求的数据
             if '?' in fields or len(fields) < 2:
                 continue
             try:
-                # 假设特征在前面，标签在最后一列
                 data.append([float(x) for x in fields[:-1]])
                 labels.append(int(fields[-1]))
             except ValueError:
-                # 如果无法转换为浮点数或整数，跳过该行
                 continue
-    # 返回值需要是两个NumPy数组
+    
     return np.array(data), np.array(labels)