I have list of list:
my_list=[['06/20', '10:02:49am', '', '-125', 'Sunday'],
['', '09:57:51am', '+116', '-126', ''],
['', '09:56:02am', '', '-123', ''],
['', '09:53:36am', '', '-124',''],
['', '09:44:22am', '+114', '-123', ''],
['', '09:41:58am', '+115', '', ''],
['', '09:39:32am', '+112', '-121', ''],
['06/19', '10:02:49am', '+200', '-125', '']]
some nested items are empty, I want to fill them with the direct previous nested list item (except the last item), required result:
[['06/20', '10:02:49am', '+116', '-125', 'sunday'],
['06/19', '09:57:51am', '+116', '-126',''],
['06/19', '09:56:02am', '+114', '-123',''],
['06/19', '09:53:36am', '+114', '-124',''],
['06/19', '09:44:22am', '+114', '-123',''],
['06/19', '09:41:58am', '+115', '-121',''],
['06/19', '10:02:49am', '+112', '-121',''],
['06/19', '10:02:49am', '+200', '-125','']]
I used pandas to do the work then return the result back to list of list, but I need a result using pure python list comprehension or numpy because I need very fast execution and I need the result as list of list not dataframe.
This is the pandas code which work very fine, but slow:
import pandas as pd
import numpy an np
df = pd.DataFrame(my_list)
df = df.replace('', np.nan)
df[0] = df[0].bfill()
df[1] = df[1].bfill()
df[2] = df[2].bfill()
df[3] = df[3].bfill()
#df
0 1 2 3 4
0 06/20 10:02:49am +116 -125 sunday
1 06/19 09:57:51am +116 -126
2 06/19 09:56:02am +114 -123
3 06/19 09:53:36am +114 -124
4 06/19 09:44:22am +114 -123
5 06/19 09:41:58am +115 -121
6 06/19 10:02:49am +112 -121
7 06/19 10:02:49am +200 -125
my_list = df.values.tolist() #work very fine, but I need an a faster alternative
