How to change the histogram bar width in displot

Question

I'm using seaborn's displot to get stacked bar charts. Here is my code:

g = sns.displot(data=df, x='collection_year', hue='WB_Income group', multiple='stack',col="ena_species",col_wrap=6,facet_kws={'sharey': False, 'sharex': False})

and here is the output

How can I increase the width of the bars?

Answer 1

• In this case, the x-axis is continuous, as is typical of a histogram.
• This issue can be resolved by setting discrete=True or setting bins= to match the number of unique values in 'year', which results in the x-ticks corresponding only to the values passed to the independent axis.
• Tested in python 3.10, pandas 1.4.3, matplotlib 3.5.1, seaborn 0.11.2

Imports & Sample DataFrame

import pandas as pd
import numpy as np
import seaborn as sns

# create sample data
np.random.seed(2022)
rows = 10000
data = {'year': np.random.choice(list(range(1985, 2025)), size=rows),
        'income': np.random.choice(['high', 'upper middle', 'lower middle', 'low'], size=rows),
        'species': np.random.choice(list(string.ascii_lowercase[:14]), size=rows)}
df = pd.DataFrame(data)

# display(df.head())
   year        income species
0  2013  upper middle       m
1  2009  lower middle       g
2  2003  lower middle       h
3  2009  upper middle       g
4  2001  lower middle       m

• The example plot, with large spaces between each bar, was only reproducible by setting bins=160, which is 4x the number of years from 1985 - 2025.
• There may be something in the 'year' data that's not apparent in the OP.

g = sns.displot(data=df, x='year', hue='income', multiple='stack', col="species", col_wrap=6, bins=160, facet_kws={'sharey': False, 'sharex': False})

• The issue can be resolved by setting bins to match the number of years in the data.

g = sns.displot(data=df, x='year', hue='income', multiple='stack', col="species", col_wrap=6, bins=40, facet_kws={'sharey': False, 'sharex': False})

• The issue can also be resolved by removing bins, and setting discrete=True.

g = sns.displot(data=df, x='year', hue='income', multiple='stack', col="species", col_wrap=6, discrete=True, facet_kws={'sharey': False, 'sharex': False})

• Setting bins to a number less than the number of unique values in 'year' will produce a histogram where each bar encompasses multiple years.

g = sns.displot(data=df, x='year', hue='income', multiple='stack', col="species", col_wrap=6, bins=20, facet_kws={'sharey': False, 'sharex': False})