Spread the love

Learn how to make beautiful GUI apps

Comprehensive, for Kivy beginners, easy to follow.

Get the book here (PDF) or on Amazon:

ebook / paperback (black and white) / paperback (full color)

SETTING AND RESETTING INDICES

In the previous part of the Pandas series we were talking about sorting indices. Today we’ll be talking about another way of rearranging multi-indexed data, index setting and resetting. Let’s start with the latter. You can use the reset_index method if you want to turn index labels into columns. Let’s use the example from the previous article to demonstrate this:

In [2]:

import numpy as np
import pandas as pd

index = pd.MultiIndex.from_product([['Columbus', 'Seattle', 'Denver', 'Dallas'], [2019, 2020]])
data = pd.Series(np.random.rand(8), index=index)
data.index.names = ['city', 'year']
data

Out[2]:

city      year
Columbus  2019    0.109463
          2020    0.634263
Seattle   2019    0.079679
          2020    0.064878
Denver    2019    0.336507
          2020    0.358905
Dallas    2019    0.845219
          2020    0.185308
dtype: float64

And now let’s reset the index. This will produce a DataFrame with the index labels ‘city’ and ‘year’ turned into column names. We’ll also pass the name for the data column:

In [5]:

data_reset = data.reset_index(name='result')
data_reset

Out[5]:

	city	year	result
0	Columbus	2019	0.109463
1	Columbus	2020	0.634263
2	Seattle	2019	0.079679
3	Seattle	2020	0.064878
4	Denver	2019	0.336507
5	Denver	2020	0.358905
6	Dallas	2019	0.845219
7	Dallas	2020	0.185308

This was resetting the index. You can also do the opposite. By setting the index you will turn a DataFrame into a multiply indexed DataFrame:

In [6]:

data_reset.set_index(['city', 'year'])

Out[6]:

		result
city	year
Columbus	2019	0.109463
Columbus	2020	0.634263
Seattle	2019	0.079679
Seattle	2020	0.064878
Denver	2019	0.336507
Denver	2020	0.358905
Dallas	2019	0.845219
Dallas	2020	0.185308