Create of multiple subsets from existing pandas dataframe

Asked Apr 13 '19 at 19:34

Active Apr 13 '19 at 19:36

Viewed 31 times

I have a pandas dataframe df as:

date       coun tk  so
2017-07-05  1.0  24  1
2017-09-04  1.0  24  2
2017-08-10  1.0  24  3
2017-10-30  1.0  24  4
2017-07-15  1.0  4   1
2017-07-31  1.0  4   1
2017-08-17  1.0  4   2
2017-05-02  1.0  4   1
2018-01-01  49.0 14  1
2017-11-27  49.0 24  1
2017-09-27  49.0 14  1
2017-10-20  49.0 24  1
2017-09-20  49.0 14  1
...

All I want to create subsets datasets from existing dataframe df for each pairs? So, all I want something like,

df1

date       coun tk  so
2017-07-05  1.0  24  1
2017-09-04  1.0  24  2
2017-08-10  1.0  24  3
2017-10-30  1.0  24  4

df2

date       coun tk  so
2017-07-15  1.0  4   1
2017-07-31  1.0  4   1
2017-08-17  1.0  4   2
2017-05-02  1.0  4   1

df3

date       coun tk  so
2018-01-01  49.0 14  1
2017-09-27  49.0 14  1
2017-09-20  49.0 14  1

df4

2017-11-27  49.0 24  1
2017-10-20  49.0 24  1

There are many other possible pairs existed in dataframe df.

edited Apr 13 '19 at 19:36

ALollz

54,844
7
56
77

asked Apr 13 '19 at 19:34

James

1

I suggest use `dfs = dict(tuple(df.groupby(['coun', 'tk'])))` – jezrael Apr 13 '19 at 19:38
@jezrael Thanks sir. How I can seperate dictionary keys, values pairs into individual dataset – James Apr 13 '19 at 19:47
Possible, but not recommended - `for i, (v, df) in enumerate(df.groupby(['coun', 'tk'])): globals()['df' + str(i)] = df.reset_index(drop=True)` – jezrael Apr 13 '19 at 19:51
Changed solution from [this](https://stackoverflow.com/a/45567689/2901002) – jezrael Apr 13 '19 at 19:52

Create of multiple subsets from existing pandas dataframe

0 Answers0