modify pandas boxplot output

Question

I made this plot in pandas, according to the documentation:

import pandas as pd
import numpy as np
import pyplot as plt

df = pd.DataFrame(np.random.rand(140, 4), columns=['A', 'B', 'C', 'D'])
df['models'] = pd.Series(np.repeat(['model1','model2', 'model3', 'model4', 'model5', 'model6', 'model7'], 20))
plt.figure()
bp = df.boxplot(by="models")

How can I modify this plot?

I want:

modify arrangement from (2,2) to (1,4)
change the labels and titles, text and font size
remove the '[models]' text

and how do I save this plot as pdf ?

Kennet Celeste · Accepted Answer · 2016-10-19T11:19:17.367

7

For the arrangement use layout
For setting x label use set_xlabel('')
For figure title use figure.subtitle()
For changing the figure size use figsize=(w,h) (inches)

note: the line np.asarray(bp).reshape(-1) is converting the layout of the subplots (2x2 for instance) to an array.

code :

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

df = pd.DataFrame(np.random.rand(140, 4), columns=['A', 'B', 'C', 'D'])
df['models'] = pd.Series(np.repeat(['model1','model2', 'model3', 'model4', 'model5', 'model6', 'model7'], 20))
bp = df.boxplot(by="models",layout=(4,1),figsize=(6,8))
[ax_tmp.set_xlabel('') for ax_tmp in np.asarray(bp).reshape(-1)]
fig = np.asarray(bp).reshape(-1)[0].get_figure()
fig.suptitle('New title here')
plt.show()

result:

edited Oct 19 '16 at 11:19

answered Oct 19 '16 at 08:33

Kennet Celeste

4,593
3
25
34

this does not work for me, i get `AttributeError: 'numpy.ndarray' object has no attribute 'set_xlabel` – spore234 Oct 19 '16 at 09:31
@spore234 comment the line `bp = df.boxplot(by="models",layout=(1,4))` and put a `print bp` before or after it, run the code and paste the result of the output here – Kennet Celeste Oct 19 '16 at 10:22
I get this from print (shortened it): [[ – spore234 Oct 19 '16 at 10:57
thanks, that works. Do you know how I can change the figsize? – spore234 Oct 19 '16 at 11:10
Let us [continue this discussion in chat](http://chat.stackoverflow.com/rooms/126184/discussion-between-spore234-and-yugi). – spore234 Oct 20 '16 at 07:01
What would be the parameter for changing the individual colors of the boxes in each subplot? – JALO - JusAnotherLivngOrganism Jun 21 '18 at 07:54

score 2 · Answer 2 · answered Oct 19 '16 at 08:32

A number of things you can do already using the boxplot function in pandas, see the documentation.

You can already modify the arrangement, and change the fontsize:

import pandas as pd
import numpy as np
import pyplot as plt

df = pd.DataFrame(np.random.rand(140, 4), columns=['A', 'B', 'C', 'D'])
df['models'] = pd.Series(np.repeat(['model1','model2', 'model3', 'model4', 'model5', 'model6', 'model7'], 20))
bp = df.boxplot(by="models", layout = (4,1), fontsize = 14)

Changing the columns the labels can be done by changing the columns labels of the dataframe itself:
```
df.columns(['E', 'F', 'G', 'H', 'models'])
```
For further customization I would use the functionality from matlotlib itself; you can take a look at the examples here.

score 0 · Answer 3 · answered Nov 14 '22 at 20:46

For those wondering how to change the individual boxplot labels (known as tick labels): model1,model2,etc they can be changed using the set_xticklables() function e.g. to rename the xtick labels to be mX:

ax=bp[-1].axes
ax.set_xticklabels([f"m{(n%7)+1}" for n in range(len(ax.get_xticklabels()))])

modify pandas boxplot output

3 Answers3

Linked