Loop through an unspecified number of pandas DataFrames

Asked Mar 30 '18 at 19:15

Active Mar 30 '18 at 19:39

Viewed 25 times

I have an n-number of pandas DataFrames each containing any number of elements, and I want produce all combinations that can be made by taking one element from each of the DataFrames.

Something like this, where I am showing three different DataFrames and I want to find the combinations that can be made only from column 'a' from each:

Input:
df1 = pd.DataFrame({'a':[1,2]},{'b':[4,4]})
df2 = pd.DataFrame({'a':[7]},{'b':[4,4]})
df3 = pd.DataFrame({'a':[1,5]},{'b':[4,4]})

Output:
1,7,1
1,7,5
2,7,1
2,7,5

I thought there might be a way to use nested for-loops but in not knowing the number of DataFrames, which may also be a large number, I'm not sure of how do that.

edited Mar 30 '18 at 19:39

asked Mar 30 '18 at 19:15

The Nightman

5,609
13
41
74

2

Are you looking for `itertools.product`? – cs95 Mar 30 '18 at 19:16
That might work but I realize in simplifying my question I may have left out some important details; I'll revise. – The Nightman Mar 30 '18 at 19:24
So it looks like `for i in product(df1['a'],df2['a'],df3['a']):` will do the trick. – The Nightman Mar 30 '18 at 19:52

Loop through an unspecified number of pandas DataFrames

0 Answers0