上面有很多很好的答案，所以我只想再加一个例子，在这种情况下，你想通过使用numpy库来指定火车和测试集的确切样本数量。

# set the random seed for the reproducibility
np.random.seed(17)

# e.g. number of samples for the training set is 1000
n_train = 1000

# shuffle the indexes
shuffled_indexes = np.arange(len(data_df))
np.random.shuffle(shuffled_indexes)

# use 'n_train' samples for training and the rest for testing
train_ids = shuffled_indexes[:n_train]
test_ids = shuffled_indexes[n_train:]

train_data = data_df.iloc[train_ids]
train_labels = labels_df.iloc[train_ids]

test_data = data_df.iloc[test_ids]
test_labels = data_df.iloc[test_ids]

我会用K-fold交叉验证。 它已被证明比train_test_split提供更好的结果。下面是一篇关于如何在sklearn中应用它的文章，来自文档本身:https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.KFold.html

不需要转换为numpy。只要用pandas df来做拆分，它就会返回一个pandas df。

from sklearn.model_selection import train_test_split

train, test = train_test_split(df, test_size=0.2)

如果你想把x和y分开

X_train, X_test, y_train, y_test = train_test_split(df[list_of_x_cols], df[y_col],test_size=0.2)

如果要分割整个df

X, y = df[list_of_x_cols], df[y_col]

要分成两个以上的类，如训练、测试和验证，可以这样做:

probs = np.random.rand(len(df))
training_mask = probs < 0.7
test_mask = (probs>=0.7) & (probs < 0.85)
validatoin_mask = probs >= 0.85


df_training = df[training_mask]
df_test = df[test_mask]
df_validation = df[validatoin_mask]

这将把大约70%的数据用于训练，15%用于测试，15%用于验证。

我会使用numpy的randn:

In [11]: df = pd.DataFrame(np.random.randn(100, 2))

In [12]: msk = np.random.rand(len(df)) < 0.8

In [13]: train = df[msk]

In [14]: test = df[~msk]

为了证明这是有效的:

In [15]: len(test)
Out[15]: 21

In [16]: len(train)
Out[16]: 79

Scikit Learn的train_test_split就是一个很好的例子。它将拆分numpy数组和数据框架。

from sklearn.model_selection import train_test_split

train, test = train_test_split(df, test_size=0.2)