Compute Grouped Data Covariance in Python

Assume, you have a dataframe and the result for calculating covariance from grouped data and corresponding column as,

Grouped data covariance is:
                  mark1       mark2
subjects
maths    mark1    25.0    12.500000
         mark2    12.5    108.333333
science  mark1    28.0    50.000000
         mark2    50.0    233.333333

Grouped data covariance between two columns:
subjects
maths    12.5
science  50.0
dtype: float64

Solution

To solve this, we will follow the steps given below −

• Define a dataframe

• Apply groupby function inside dataframe subjects column

df.groupby('subjects')

• Apply covariance function to grouped data and store insied group_data,

group_data = df.groupby('subjects').cov()

• Apply lambda function for mark1 and mark2 columns with groupby records from the subjects column. It is defined below,

df.groupby('subjects').apply(lambda x: x['mark1'].cov(x['mark2']

Example

Let’s see the below code to get a better understanding −

import pandas as pd
df =
pd.DataFrame({'subjects':['maths','maths','maths','science','science','science'],
               'mark1':[80,90,85,95,93,85],
               'mark2':[85,90,70,75,95,65]})
print("DataFrame is:\n",df)
group_data = df.groupby('subjects').cov()
print("Grouped data covariance is:\n", group_data)
result = df.groupby('subjects').apply(lambda x: x['mark1'].cov(x['mark2']))
print("Grouped data covariance between two columns:\n",result)

Output

DataFrame is:
   subjects mark1 mark2
0    maths   80    85
1    maths   90    90
2    maths   85    70
3    science 95    75
4    science 93    95
5    science 85    65
Grouped data covariance is:
               mark1    mark2
subjects
maths    mark1 25.0    12.500000
         mark2 12.5    108.333333
science  mark1 28.0    50.000000
         mark2 50.0    233.333333

Grouped data covariance between two columns:
subjects
maths    12.5
science  50.0
dtype: float64