CODEWITHSUNDEEP
pythonpandas
Emm
Emm

Reputation: 2507

Conditionally summing multiple columns

I would like to sum certain rows based on a condition in a different row.

So I have a columns for points 

{'secondBoxer1': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0},
 'secondBoxer2': {0: 0.0, 1: 0.0, 2: 10.0, 3: 0.0, 4: 0.0},
 'secondBoxer3': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0},
 'secondBoxer4': {0: 15.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0},
 'secondBoxer5': {0: 15.0, 1: 53.57142857142857, 2: 0.0, 3: 0.0, 4: 0.0},
 'secondBoxer6': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0},
 'secondBoxer7': {0: 0.0, 1: 0.0, 2: 0.0, 3: 50.0, 4: 0.0},
 'secondBoxer8': {0: 0.0, 1: 0.0, 2: 0.0, 3: 37.142857142857146, 4: 0.0}}

and a column with the outcome of each fight

{'outcome1': {0: 'win ', 1: 'win ', 2: 'win ', 3: 'draw ', 4: 'win '},
 'outcome2': {0: 'win ', 1: 'win ', 2: 'win ', 3: 'win ', 4: 'win '},
 'outcome3': {0: 'win ', 1: 'win ', 2: 'win ', 3: 'win ', 4: 'scheduled '},
 'outcome4': {0: 'win ', 1: 'win ', 2: 'nan', 3: 'loss ', 4: 'nan'},
 'outcome5': {0: 'win ', 1: 'draw ', 2: 'nan', 3: 'win ', 4: 'nan'},
 'outcome6': {0: 'nan', 1: 'nan', 2: 'nan', 3: 'loss ', 4: 'nan'},
 'outcome7': {0: 'nan', 1: 'nan', 2: 'nan', 3: 'loss ', 4: 'nan'},
 'outcome8': {0: 'nan', 1: 'nan', 2: 'nan', 3: 'win ', 4: 'nan'}}

I would like to sum the points in the first columns (points columns) in cases where the outcome is equals to a win.

I have written this code, where opp_names is the list of columns with the points and outcome_cols is a list of columns with the outcomes

data[opp_names].sum(axis=1).where(data[outcome_cols] == 'win')

The problem with the output from this code is that it returns a total sum of points that is not conditional

Upvotes: 0

Views: 42

Answers (1)

BENY
BENY

Reputation: 323226

In your case we use mask :d is your first dict , d1 is your 2nd dict

pd.DataFrame(d).mask(pd.DataFrame(d1).ne('win ').to_numpy()).sum(1)
Out[164]: 
0    30.000000
1     0.000000
2    10.000000
3    37.142857
4     0.000000
dtype: float64

Upvotes: 1

Related Questions