从 plot 中删除默认补丁

Question

I want to export a plot that contains no "unnecessary" patches. The following code exports a figure that has invisible patches with zero alpha:

plt.title("Test")
plt.plot()
plt.gca().patch.set_alpha(0.)
plt.gcf().patch.set_alpha(0.)
plt.savefig('test.svg')

Although these patches are not visible, their geometry hinders Inkscape from properly resizing the page to its content. Would these "unnecessary" patches be removed the grey area could be cropped using eg Inkscape:

How do I either plot without the patches or remove the figure's and axis' patches programmatically?

Answer 1

SVG is a text file. Use a text editor and delete these parts:

   <g id="patch_3">
    <path d="M 54 252 
L 54 34.56 
" style="fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;"/>
   </g>
   <g id="patch_4">
    <path d="M 388.8 252 
L 388.8 34.56 
" style="fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;"/>
   </g>
   <g id="patch_5">
    <path d="M 54 252 
L 388.8 252 
" style="fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;"/>
   </g>
   <g id="patch_6">
    <path d="M 54 34.56 
L 388.8 34.56 
" style="fill:none;stroke:#000000;stroke-linecap:square;stroke-linejoin:miter;stroke-width:0.8;"/>
   </g>

The details can be different, but you can figure out with fill:none in the style.

EDIT 1 SVG file can be rendered and edit with Inkscape. With its XML Editor, it is easy to find all the unneeded elements and delete them in XML Editor.

Answer 2

This seems to do the trick for this plain MWE. I hope it fits your needs.

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

fig, ax = plt.subplots()

plt.title("Test")
ax.plot()

plt.savefig('test.svg', bbox_inches =  TransformedBbox(Bbox(ax.bbox.get_points()),transform=fig.dpi_scale_trans.inverted()) )

The strategy is to get the ax bbox and then use it as an argument to savefig (it requires transformation to figure inches).

More in here:

https://matplotlib.org/3.1.1/api/transformations.html#matplotlib.transforms.TransformedBbox

https://matplotlib.org/tutorials/advanced/transforms_tutorial.html

EDIT

A cleaner solution:

import matplotlib.pyplot as plt

fig, ax = plt.subplots()

plt.title("Test")
ax.plot()

fig.patch.set_visible(False)

plt.savefig('test.svg',)

The best solution now depends on your specific case.