Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

svg backend truncates output size to integer, which it doesn't need to (and pdf backend doesn't) #9815

anntzer opened this issue Nov 19, 2017 · 1 comment


None yet
3 participants
Copy link

commented Nov 19, 2017

Bug report

Bug summary

For vector backends, it should be acceptable for the output canvas to have a noninteger size. Indeed, the pdf backend will output canvas with noninteger sizes "as is". On the other hand, the svg backend truncates the output to the integer below, which it doesn't need to do.

Code for reproduction

gcf(); savefig("/tmp/foo.pdf"); savefig("/tmp/foo.svg")

and examine foo.pdf and foo.svg with a plain text viewer.

Actual outcome

foo.pdf contains the line

/MediaBox [ 0 0 460.8 345.6 ] /Parent 2 0 R /Resources 8 0 R /Type /Page >>

(460.8 is 6.4 inches at 72dpi, 345.6 is 4.8 inches at 72 dpi).

foo.svg contains the line

<svg height="345pt" version="1.1" viewBox="0 0 460 345" width="460pt" xmlns="" xmlns:xlink="">

so the SVG was truncated. Note that interestingly, the SVG also contains e.g.

   <path d="M 0 345.6 
L 460.8 345.6 
L 460.8 0 
L 0 0 
" style="fill:#ffffff;"/>

i.e. some paths that overflow the SVG canvas.

I believe it is just as valid for the SVG output to use noninteger sizes (at least, manually editing the file doesn't make inkscape complain).

Expected outcome

SVG backend should not crop the output size.

Matplotlib version

  • Operating system: N/A
  • Matplotlib version: 2.1/master
  • Matplotlib backend (print(matplotlib.get_backend())): SVG
  • Python version: N/A
  • Jupyter version (if applicable): N/A
  • Other libraries: N/A

@anntzer anntzer added the backend/svg label Nov 19, 2017

@tacaswell tacaswell added this to the v2.2 milestone Nov 19, 2017


This comment has been minimized.

Copy link

commented Nov 19, 2017

It appears have been this way since at least 2005 but see no reason for it to be so.

tacaswell added a commit to tacaswell/matplotlib that referenced this issue Nov 19, 2017

@QuLogic QuLogic modified the milestones: needs sorting, v2.2.0 Feb 12, 2018

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
You can’t perform that action at this time.