Numerical Computing with Python

How to do it...

1. Import the employee data and select the DEPARTMENT and BASE_SALARY columns in a new DataFrame:

>>> employee = pd.read_csv('data/employee.csv')>>> dept_sal = employee[['DEPARTMENT', 'BASE_SALARY']]

2. Sort this smaller DataFrame by salary within each department:

>>> dept_sal = dept_sal.sort_values(['DEPARTMENT', 'BASE_SALARY'],                                       ascending=[True, False])

3. Use the drop_duplicates method to keep the first row of each DEPARTMENT:

>>> max_dept_sal = dept_sal.drop_duplicates(subset='DEPARTMENT')>>> max_dept_sal.head()

4. Put the DEPARTMENT column into the index for each DataFrames:

>>> max_dept_sal = max_dept_sal.set_index('DEPARTMENT') ...