2

So I'm trying to do a bar plot on a data where x is the username (string ) and each x is long enough to overlap each other, so I have to rotate the x label. No problem there. However, when exporting the plot results, the x label on the exported image is cropped. I tried using plt.tight_layout() and worked, but it change the look of the plot. The code is similar to this


from matplotlib import pyplot as plt

x= ['abc', 'ronaldo', 'melon_killer_123456']
y= [1, 2, 3]

plt.bar(x, y)

plt.xticks(rotation = 90) 
plt.savefig('a.png')

plt.show()

Exported image:

wrong_image

I want it to look like this (Got this by using jupyter notebook and manually save output image):

correct_image

So how to do that?

Yusuf Syam
  • 701
  • 1
  • 4
  • 18
  • when I ran your supplied code I got an error, and you forgot to import matplotlib as plt – Hezekiah Bodine Jul 12 '22 at 15:46
  • Similar to this topic https://stackoverflow.com/questions/61192115/why-doesnt-matplotlib-save-the-whole-figure-by-default Adding plt.savefig(..,bbox_inches='tight') solved the problem for me (giving the same output as Jupyter notebooks) – Arno Deceuninck May 18 '23 at 08:42

2 Answers2

2

You can play around with the rcParams size settings and the plt.subplots_adjust settings until you get your desired image.

import matplotlib.pyplot as plt
x= ['abc', 'ronaldo', 'melon_killer_123456']
y= [1, 2, 3]

plt.rcParams["figure.figsize"] = (5,10)
plt.bar(x, y)
plt.xticks(rotation = 90) 
plt.subplots_adjust(top=0.925, 
                    bottom=0.20, 
                    left=0.07, 
                    right=0.90, 
                    hspace=0.01, 
                    wspace=0.01)
plt.savefig('a.png')
plt.show()
Michael S.
  • 3,050
  • 4
  • 19
  • 34
2

So i got another answer by assign a Bbox instance to bbox_inches parameter on plt.savefig() function. With this Bbox we can define xmin, ymin xmaxand ymax of our saved figure (Bbox([[xmin, ymin],[xmax, ymax]])). The default value is xmin=0, ymin=0 xmax=figure_widthand ymax=figure_height

If the problem is on the bottom of the saved figure, we need just configure (lower) the ymin


import matplotlib.pyplot as plt
from matplotlib.transforms import Bbox

x= ['abc', 'ronaldo', 'melon_killer_123456']
y= [1, 2, 3]

plt.bar(x, y)
plt.xticks(rotation = 90) 
plt.savefig('a.png', bbox_inches=Bbox([[0,-2],fig.get_size_inches()]))
plt.show()

P.s fig.get_size_inches() give us a list of the fig size

jyyj yj
  • 29
  • 5