pandas Group by and convert from row of list to column for a large file (No transpose)

Asked Jun 26 '18 at 11:20

Active Jun 26 '18 at 12:04

Viewed 41 times

After using pandas Group by data frame becomes

model  item_list
a      [l1,l5,l3,l4,l7,l9,l10]
b      [l2,l6,l1,l5]
c      [l1,l3]

And the expected output is

model  item_list
    a      l1
    a      l5
    a      l4
    a      l7
    a      l9
    a      l10
    b      l2
    b      l6
    b      l1
    b      l5
    c      l1
    c      l3

Is there any approach to get it done with less computation as the actual data is very large?

Please note the original dataframe row item_list consist of []

edited Jun 26 '18 at 11:57

asked Jun 26 '18 at 11:20

Jahnab Kumar Deka

Check [maxu](https://stackoverflow.com/a/40449726/2901002) answer. – jezrael Jun 26 '18 at 11:22
added new 2 solutions to dupes, maybe better here. – jezrael Jun 26 '18 at 11:25
is there any option while doing in groupby itself? @jezrael – Jahnab Kumar Deka Jun 26 '18 at 11:45
But why need it? I think it is not necessary. Can you explain more? – jezrael Jun 26 '18 at 11:47
yes actually after doing groupby result is in the form of array. So other solution won't work directly. i have converted it to pandas data frame after groupby but the item list is in the form of array in each row, eg: [l1,l5,l3,l4,l7,l9,l10] @jezrael – Jahnab Kumar Deka Jun 26 '18 at 11:51
Can you edit question? – jezrael Jun 26 '18 at 11:53
Edited the question. Sorry for not clearly defining the problem earlier. @jezrael – Jahnab Kumar Deka Jun 26 '18 at 11:56
hmmm, then use `Maxu` solution, first dupe. – jezrael Jun 26 '18 at 11:56
have tried that too and getting error as explode() got an unexpected keyword argument 'fill_value' @jezrael – Jahnab Kumar Deka Jun 26 '18 at 12:02
Added new dupe, please check it. – jezrael Jun 26 '18 at 12:04

pandas Group by and convert from row of list to column for a large file (No transpose)

0 Answers0