按子列将 Dataframe 拆分为多个 Dataframe

Question

I could not find an elegant way to split this table:

Open              Close                High 
 stock1 stock2     stock1 stock2        stock1 stock2
  ..     ..         ..      ..           ..     ..

Where stocks are sub columns, into multiple tables each for a stock, so first table is

df_stock1

   Open Close High

so a table/dataframe for each stock ( could be > 2 ) then return an array of those dataframes .

Answer 1

You could do something like this. Here's a fictiv dataframe:

import pandas as pd
import numpy as np
np.random.seed(123)
df = pd.DataFrame(np.random.randint(1, 10, (6, 6)),
                  columns=pd.MultiIndex.from_product([['Open', 'Close', 'High'],
                                                      ['Stock1', 'Stock2']]))

which prints:

Open         Close          High       
  Stock1 Stock2 Stock1 Stock2 Stock1 Stock2
0      3      3      7      2      4      7
1      2      1      2      1      1      4
2      5      1      1      5      2      8
3      4      3      5      8      3      5
4      9      1      8      4      5      7
5      2      6      7      3      2      9

Then to split your multi-index df, do the following, using groupby and DataFrame.xs:

Split = {idx: df_sub.xs(idx, level=1, axis=1) for idx,df_sub in df.groupby(level=1, axis=1)}
print(Split['Stock1'])

where df_sub refers to the first-level of the multiindex df. This givess:

 Open  Close  High
0     3      7     4
1     2      2     1
2     5      1     2
3     4      5     3
4     9      8     5
5     2      7     2

and

Split2 = {idx: df_sub.xs(idx, level=1, axis=1) for idx,df_sub in df.groupby(level=1, axis=1)}
print(Split2['Stock2'])

prints:

 Open  Close  High
0     3      2     7
1     1      1     4
2     1      5     8
3     3      8     5
4     1      4     7
5     6      3     9

Other levels

Similarily, if you want all the Open stocks, you can do this:

Split_open = {idx: df_sub.xs(idx, level=0, axis=1) for idx,df_sub in df.groupby(level=0, axis=1)}
print(Split_open['Open'])

which returns:

   Stock1  Stock2
0       3       3
1       2       1
2       5       1
3       4       3
4       9       1
5       2       6