CODEWITHSUNDEEP
pythonpandaspandas-groupby
user1254513
user1254513

Reputation: 71

Python Dataframe Conditional Sum

I have a dataframe incomeData with Country, Region and Income. I am attempting to use aggregation to return mean, min, max and counts. I want to be able to count the countries where the income is greater than 100.

raw_data = {'Country': ['A', 'B', 'C', 'D', 'E'],
            'Region': ['X', 'X', 'X', 'Y', 'Y'],
            'Income': [100, 200, 300, 100, 200]
           }
incomeData = pd.DataFrame(raw_data, columns = ['Country', 'Region', 'Income'])
regionGroup = incomeData.groupby(['Region'], as_index=False)
groupCount = lambda x: x.count()
#CountHighIncome = ?
aggregations = {
    'Country': {groupCount
    },
    'Income': {'min', 'max', 'mean', 'median' #, CountHighIncome
    }
}
incomeSummary = regionGroup.agg(aggregations)
incomeSummary

   Region  Country Income

            lambda> median  max mean min CountHighIncome

0   X       3       200     300 200  100 2

1   Y       2       150     200 150  100 1

Please let me know if the lambda approach to count countries within a region can be extended to count countries within a region where the income is greater than 100. Or if there is an alternative better approach to this problem.

Many Thanks in Advance.

Upvotes: 1

Views: 645

Answers (1)

jezrael
jezrael

Reputation: 862641

You can use custom function with lambda with sum condition, Trues are count like 1, also for Country is removed lambda function and is used count only:

CountHighIncome = lambda x: (x > 100).sum()
aggregations = {
    'Country': {'count'
    },
    'Income': {'min', 'max', 'mean', 'median',  CountHighIncome
    }
}
incomeSummary = regionGroup.agg(aggregations)
print (incomeSummary)
  Region Income                           Country
            max  min <lambda> mean median   count
0      X    300  100        2  200    200       3
1      Y    200  100        1  150    150       2

Upvotes: 1

Related Questions