我尝试使用以下代码对6个数据集进行数据分割:

from sklearn.datasets import make_blobsfrom sklearn.model_selection import train_test_splitdf4, df3, df1, df2, df5, df6 = make_blobs(n_samples=2000000)df4_train, df4_test, df3_train, df3_test, df1_train, df1_test, df2_test, df2_train, df5_train, df5_test, df6_train, df6_test = train_test_split(df4, df3, test_size=0.30, random_state=10000, shuffle =True)data_frames = [list() for x in range(6)]

然而，出现了以下错误:

File "C:\Users\abd77\.spyder-py3\test_subject_1\data split.py", line 12, in <module>    df4, df3, df1, df2, df5, df6 = make_blobs(n_samples=2000000)ValueError: not enough values to unpack (expected 6, got 2)

回答：

问题出现在你创建blob数据的第4行。Scikit-learn的make_blobs方法返回一个X（输入）和一个y值（输出）。你不能告诉Python期望6个不同的输出值，而这个方法只返回2个。

https://scikit-learn.org/stable/modules/generated/sklearn.datasets.make_blobs.html

要创建6个不同的数据集，你需要调用make_blobs方法6次

dfs = [make_blobs(n_samples=2000000) for i in range(6)]dfs_split = [train_test_split(X,y) for X,y in dfs]

然后你可以使用以下方式访问每个X_train, X_test, y_train, y_test:

##对于第一个数据集X_train_df1 = dfs_split[0][0]X_test_df1 = dfs_split[0][1]y_train_df1 = dfs_split[0][2]y_test_df1 = dfs_split[0][3]