CODEWITHSUNDEEP
pythonpandasdataframepandas-groupby
JakeP
JakeP

Reputation: 107

groupby multiple values in a column

I have a subset of a dataframe here:

data = {'Name': ['ch1', 'ch2', 'ch3', 'ch4', 'ch5', 'ch6'],
        'Time': [1,2,3,4,5,6],
        'Week' : [1, 2, 3, 2, 3, 2]
            }

dfx = pd.DataFrame(data)

I need to sum up all the times for each week so Week 1 time is 1, Week 2 time is 2+4+6, and Week 3 is 3+5. I also need it to look through the 'Week' column and find all the different weeks, so for this example there are 3 but for another dataframe it could be 2 or 4.

End result is look through a column in a dataframe, find the unique values (1,2,3,...n), groupby be each of those values into rows and sum up the time for each of those values.

I have tried a handful of ways but nothing is really working how I would like. I appreciate any help or ideas.

Expected Output:

                                 Sum
Week 1: 1                         1
Week 2: 2  4  6                   12
Week 3: 3  5                      8

The output can be either individual dataframes of the data or one dataframe that has all three rows with all the numbers and the sum at the end.

Upvotes: 0

Views: 822

Answers (2)

user7864386
user7864386

Reputation:

You can groupby "Week", select column "Time", and you can pass multiple functions (such as list constructor and sum) to Groupby.agg to do the things you want:

out = dfx.groupby('Week')['Time'].agg(Times=list, Total=sum)

Output:

          Times  Total
Week                  
1           [1]      1
2     [2, 4, 6]     12
3        [3, 5]      8

Upvotes: 1

Shivang Gupta
Shivang Gupta

Reputation: 347

    import pandas as pd
data = {'Name': ['ch1', 'ch2', 'ch3', 'ch4', 'ch5', 'ch6'],
        'Time': [1,2,3,4,5,6],
        'Week' : [1, 2, 3, 2, 3, 2]
            }

dfx = pd.DataFrame(data)
dfx = dfx.groupby('Week')['Time'].sum()
print(dfx)

output:

Week
1     1
2    12
3     8

Upvotes: 1

Related Questions