bzog.md

lang.py.module.pandas.index.sorting

Sorting base on the row index or column labels

Synopsis

  index_sorted_df = df.sort_index()

Overview

Unlike sort_values this will perform sorting based on the row index or column names and not by their values

When you perform sorting, filtering or drop rows the index label get messed up. You can sort based on this index numbers or labels with sort_index()

Rember: By default the result of sort_index returns a copy of the original dataframe after being sort, the original dataframe is unafected

Cookbook

Sort rows based on index

After applying sorting the original index labels may have become out of order. You can sort them back with sort_index(). By default it applies ascending sorting

  sorted_df = df.sort_values(by=["make", "model"])
  sorted_df.sort_index(ascending=True)

Sort the columns of your dataframe

You can sort your colummns with the axis="columns" argument, so that they will be in ascending or descending order based on their label names.

You can also use the columns name or the number 1

  df.sort_index(axis="columns", ascending=False)
  df.sort_index(axis=1, ascending=False)    # Same result as before