What I need is to calculate a number of actions by every month in the period from '2014-03' to '2019-01' for each module from start.
month_range = ['2014-03', '2014-04', '2014-05',
'2014-06', '2014-07', '2018-10',
'2018-11', '2018-12', '2019-01']
df = pd.DataFrame({'action':['done','in_work', 'done',
'fail','in_work', 'done'],
'module':['B1','B2','B1',
'B3','B1','B2'],
'start': ['2014-06','2014-07','2014-07',
'2014-10','2018-09','2018-12'],
'finish':['2014-06', NaT, '2018-10',
'2014-10', NaT, '2019-02']
},
columns = ['action','module',
'start', 'finish'])
A huge limitation. One action can be made from one month to another and be counted by every month in a date range.
For example, if module is still 'in work' & action has no ending (NaT) it should be counted by every month from the begining to max date in range.
Something in form like this is expected after calculating:
B1 B2 B3
1 2014-03 0 0 0
2 2014-04 0 0 0
3 2014-05 0 0 0
4 2014-06 1 0 0
5 2014-07 1 0 0
If someone knows any solution?