CODEWITHSUNDEEP
pythonpandas
Dinosaurius
Dinosaurius

Reputation: 8628

How to add a condition to grouping?

I have the following dataframe:

df = 
GROUP   TOTAL_SERVICE_TIME  TOTAL_WAIT_TIME IS_EVALUATED    IS_NEGATIVE_GRADE
AAA     19                  60              0               0
AAA     248                 84              1               0
AAA     135                 62              1               1
BBB     97                  36              1               1
BBB     395                 117             0               0

I am grouping the data as follows (by GROUP and TOTAL_WAIT_TIME):

funcs = {
    'TOTAL_SERVICE_TIME': {'TOTAL_SERVICE_TIME':'mean'},
    'IS_EVALUATED' : {'IS_EVALUATED':'size'},
    'IS_NEGATIVE_GRADE' : {'IS_NEGATIVE_GRADE':'size'},
}

fresult = result.groupby(['GROUP','TOTAL_WAIT_TIME']).agg(funcs)
fresult.columns = fresult.columns.droplevel(0)
fresult = fresult.reset_index()
fresult

The problem is that IS_EVALUATED and IS_NEGATIVE_GRADE are calculated incorrectly. I want to count only the values of 1 in these columns, but not all the rows.

Upvotes: 1

Views: 37

Answers (1)

MaxU - stand with Ukraine
MaxU - stand with Ukraine

Reputation: 210912

In [202]: funcs = {
     ...:     'TOTAL_SERVICE_TIME': 'mean',
     ...:     'IS_EVALUATED' : 'sum',
     ...:     'IS_NEGATIVE_GRADE' : 'sum',
     ...: }
     ...:
     ...: fresult = result.groupby(['GROUP','TOTAL_WAIT_TIME'], as_index=False).agg(funcs)
     ...:

In [203]: fresult
Out[203]:
  GROUP  TOTAL_WAIT_TIME  IS_EVALUATED  IS_NEGATIVE_GRADE  TOTAL_SERVICE_TIME
0   AAA               60             0                  0                  19
1   AAA               62             1                  1                 135
2   AAA               84             1                  0                 248
3   BBB               36             1                  1                  97
4   BBB              117             0                  0                 395

Upvotes: 1

Related Questions