在GridSearchCV中明确指定测试/训练集

正如@MaxU所说，最好让GridSearchCV处理拆分，但是如果您要按照问题中的设置执行拆分，则可以使用

PredefinedSplit

因此，您需要对代码进行以下更改。

# Here X_test, y_test is the untouched data# Validation data (X_val, y_val) is currently inside X_train, which will be split using PredefinedSplit inside GridSearchCVX_train, X_test = np.array_split(X, [50])y_train, y_test = np.array_split(y, [50])# The indices which have the value -1 will be kept in train.train_indices = np.full((35,), -1, dtype=int)# The indices which have zero or positive values, will be kept in testtest_indices = np.full((15,), 0, dtype=int)test_fold = np.append(train_indices, test_indices)print(test_fold)# OUTPUT: array([-1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,       -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,       -1,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0])from sklearn.model_selection import PredefinedSplitps = PredefinedSplit(test_fold)# Check how many splits will be done, based on test_foldps.get_n_splits()# OUTPUT: 1for train_index, test_index in ps.split():    print("TRAIN:", train_index, "TEST:", test_index)# OUTPUT: ('TRAIN:', array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16,   17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33,   34]),  'TEST:', array([35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49]))# And now, send this `ps` to cv param in GridSearchCVfrom sklearn.model_selection import GridSearchCVgrid_search = GridSearchCV(Ridge(random_state=444), param_grid, cv=ps)# Here, send the X_train and y_traingrid_search.fit(X_train, y_train)

发送到X_train和y_train的代码

fit()

希望这能清除工作。