Increasing space between bins in seaborn distplot

Question

So I have this, probably, simple question. I created a histogram from data out of an excel file with seaborn. Forbetter visualization, I would like to have some space between the bars/bins. Is that possible?

My code looks as followed

import pandas as pd
import matplotlib as mpl
import matplotlib.pyplot as plt
import numpy as np
import seaborn as sns

%matplotlib inline
from IPython.display import set_matplotlib_formats
set_matplotlib_formats('svg', 'pdf')


df = pd.read_excel('test.xlsx')
sns.set_style("white")
#sns.set_style("dark")
plt.figure(figsize=(12,10))
plt.xlabel('a', fontsize=18)
plt.ylabel('test2', fontsize=18)

plt.title ('tests ^2', fontsize=22)


ax = sns.distplot(st,bins=34, kde=False, hist_kws={'range':(0,1), 'edgecolor':'black', 'alpha':1.0}, axlabel='test1')

A second question though a bit off topic would be, how I get the exponent in the title of the chart to actually be uplifted?

Thanks!

@jojo: Thanks for the link. I was hoping to get something more straight forward. Otherwise I have to export the graphs as svg and correct it with inkscape etc. — Jul, Nov 07 '17 at 18:03
Ok, dump question, I still need a latex distributor for it to work? — Jul, Nov 07 '17 at 18:45

score 16 · Accepted Answer · answered Nov 06 '17 at 21:26

16

The matplotlib hist function has an argument rwidth

rwidth : scalar or None, optional
The relative width of the bars as a fraction of the bin width.

You can use this inside the distplot via the hist_kws argument.

import matplotlib.pyplot as plt
import numpy as np
import seaborn as sns

x = np.random.normal(0.5,0.2,1600)

ax = sns.distplot(x,bins=34, kde=False, 
                  hist_kws={"rwidth":0.75,'edgecolor':'black', 'alpha':1.0})

plt.show()

answered Nov 06 '17 at 21:26

ImportanceOfBeingErnest

321,279
53
665
712

1

Doesn't work with seaborn >= 0.11 any more, unfortunately. `distplot()` became `displot()` or `histplot()`, and neither passes on the rwidth parameter. – MERose Dec 21 '20 at 16:09
1

2022: The `hist_kws` argument **is** available for `sns.distplot()`, but not for `sns.histplot()`. – PatrickT Dec 28 '21 at 08:47

miro · Answer 2 · 2021-12-29T19:23:05.920

9

for Seaborn >= 0.11, use shrink parameter. It scales the width of each bar relative to the binwidth by this parameter. The rest will be empty space.

Documentation: https://seaborn.pydata.org/generated/seaborn.histplot.html

edit: OP was originally asking about sns.distplot(), however, it is deprecated in favor of sns.histplot or sns.displot() in the current version >=0.11. Since OP is generating a histogram, both histplot and displot in hist mode will take shrink

edited Dec 29 '21 at 19:23

answered Jan 26 '21 at 14:01

miro

735
8
16

The OP is asking about `sns.displot()`, while `sns.histplot()` takes different arguments. – PatrickT Dec 28 '21 at 08:48
1

@PatrickT thanks for noticing, I have improved the answer. – miro Dec 29 '21 at 19:23
upvote from me. :-) – PatrickT Dec 30 '21 at 01:33

J4FFLE · Answer 3 · 2021-05-27T20:39:02.803

After posting my answer, I realized I answered the opposite of what was being asked. I found this question while trying to figure out how to remove the space between bars. I almost deleted my answer, but in case anyone else stumbles on this question and is trying to remove the space between bars in seaborn's histplot, I'll leave it for now.

Thanks to @miro for Seaborn's updated documentation, I found that element='step' worked for me. Depending on exactly what you want, element='poly' may be what you are after.

My implementation with 'step':

fig,axs = plt.subplots(4,2,figsize=(10,10))
i,j = 0,0
for col in cols:
    sns.histplot(df[col],ax=axs[i,j],bins=100,element='step')
    axs[i,j].set(title="",ylabel='Frequency',xlabel=labels[col])
    i+=1
    if i == 4: 
        i = 0
        j+=1

My implementation with 'poly':

fig,axs = plt.subplots(4,2,figsize=(10,10))
i,j = 0,0
for col in cols:
    sns.histplot(df[col],ax=axs[i,j],bins=100,element='poly')
    axs[i,j].set(title="",ylabel='Frequency',xlabel=labels[col])
    i+=1
    if i == 4: 
        i = 0
        j+=1

The OP is asking about `sns.distplot()` not `sns.histplot()`. As it turns out, they are quite different. — PatrickT, Dec 28 '21 at 08:50

Increasing space between bins in seaborn distplot

3 Answers3