2.6 Sorting

Like native Python lists and dictionaries, DataFrames can also by one or more columns. We do this using either the or the method.

As you might have guessed, sort_index() will sort whatever the row index is set to by default. Remember that if a labeled index is not set, then a numeric RangeIndex is autogenerated and used. However, if you set a labeled row index, and if that index is not sorted alphabetically, then sort_index() will take care of that. You can also reverse sort:

 
      import pandas as pd
      df = pd.DataFrame({'age':[29, 55, 65, 18], 'gender':['male', 'female', 'male', 'female'], 'hr1':[98.1, 78, 65, 64], 'hr2':[110, 120, 129, 141], 'hr3':[76, 87, 77, 59]}, index=['p1', 'p2', 'p3', 'p4'])
      df.sort_index(ascending=False, inplace=True)
      df

You can use this method for column indexes as well:

 
      import pandas as pd
      df = pd.DataFrame({'age':[29, 55, 65, 18], 'gender':['male', 'female', 'male', 'female'], 'hr1':[98.1, 78, 65, 64], 'hr2':[110, 120, 129, 141], 'hr3':[76, 87, 77, 59]}, index=['p1', 'p2', 'p3', 'p4'])
      df.sort_index(ascending=False, inplace=True, axis=1)
      df

You can sort by the values in a row or column as well using sort_values:

 
      import pandas as pd
      df = pd.DataFrame({'age':[29, 55, 65, 18], 'gender':['male', 'female', 'male', 'female'], 'hr1':[98.1, 78, 65, 64], 'hr2':[110, 120, 129, 141], 'hr3':[76, 87, 77, 59]}, index=['p1', 'p2', 'p3', 'p4'])
      df.sort_values(by=['age'], inplace=True)
      df

In addition, you can sort by multiple levels:

 
      import pandas as pd
      df = pd.DataFrame({'age':[29, 55, 65, 18], 'gender':['male', 'female', 'male', 'female'], 'hr1':[98.1, 78, 65, 64], 'hr2':[110, 120, 129, 141], 'hr3':[76, 87, 77, 59]}, index=['p1', 'p2', 'p3', 'p4'])
      df.sort_values(by=['gender', 'age'], inplace=True)
      df
Previous Next