Seaborn barplot/catplot with grouped bars where a columns are group elements

Question

I'm not sure if this is possible with Seaborn or if I have to transform my data first. I want a barplot (or catplot) with grouped bars. The keypoint is that the elements in the groups are corosponding to column names and not to values.

That is the data

   Year  Gin  Beer
0  2016   84    78
1  2017   75    30
2  2018   42    47
3  2019   25    58
4  2020   51    90
5  2021   40    50

And that is my desired output:

From the Seaborn logic "combined" with my logic I would like to do something like this where I specify the column names in y= argument.

sns.catplot(data=a, x='Year', y=['Gin', 'Beer'])

Can I realize it without touching the data or do I have to transform the data somehow?

This is the MWE

#!/usr/bin/env python3
import random as rd
import pandas as pd
from matplotlib import pyplot as plt
import seaborn as sns
sns.set_theme()

rd.seed(0)

a = pd.DataFrame({
    'Year': [2016, 2017, 2018, 2019, 2020, 2021],
    'Gin': rd.choices(range(100), k=6),
    'Beer': rd.choices(range(100), k=6)
})
print(a)

# sns.catplot(data=a, x='Year', y=['Gin', 'Beer'])
# plt.show()

Your data needs `melt`ed first. Lots of seaborn examples on SO showing this. — BigBen, Sep 15 '22 at 19:20
In order to have that you will need to reshape and to have Gin / Beer in the same column, and in another column there respective values. That way, you can include parameter ```hue```, which will give you what you need. — sophocles, Sep 15 '22 at 19:22
Otherwise `a = a.melt(id_vars='Year')` and then `ax = sns.catplot(kind='bar', data=a, x='Year', y='value', hue='variable')` — Trenton McKinney, Sep 15 '22 at 19:28
My brain fart: just `ax = a.plot(kind='bar', x='Year', rot=0)`. No need to set index. — Trenton McKinney, Sep 15 '22 at 19:34

Seaborn barplot/catplot with grouped bars where a columns are group elements

0 Answers0