Piecewise plot not smooth on Jupyter

[Simon Willerton]

Hi!

There seems to be a bug in the plotting in the Jupyter notebook.  The following piecewise defined function should be smooth at r=1 but the plot is not; the first part of the plot seems to be lifted slightly.  I noticed this in a more complicated example but this is a reasonably minimal case.

r = var('r')
plot(2*r - r^2, r, 0, 1) + plot(1, r, 1, 2)

This bug doesn't seem to occur in CoCalc .sagews notebook but does occur in SageMath Kernels 7.6, 8.0, 8.1 in a CoCalc Jupyter notebook and offline on my Mac OS X Sierra in a Jupyter notebook.

Cheers,

Simon.

[William Stein]

It could be an issue with svg versus png output.

Cheers,

Simon.

--
You received this message because you are subscribed to the Google Groups "sage-support" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sage-support...@googlegroups.com.
To post to this group, send email to sage-s...@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-support.
For more options, visit https://groups.google.com/d/optout.

--

-- William Stein

[Simon Willerton]

Indeed, that seems to be the problem.  Comparing the output from the following code gives the svg as smooth but the png not smooth.

p = plot(2*r- r^2, r, 0,1) + plot(1, r, 1, 2)
p.save("plot.svg")
p.save("plot.png")
 

[Simon Willerton]

Following William's hint, I came up with a workaround.  The first solution I found at https://stackoverflow.com/a/45664729 didn't work on .sagews files, so combining that with https://stackoverflow.com/a/44100805, I ended up with the following.  Do tell me if there's a better way!

from IPython.display import SVG, display

def svg_show(pic, *args, **kwds):
    """
    Display the plot 'pic' using svg (rather than png).  If we are in Jupyter (ie. IPython)
    we need to do this explicitly.  Otherwise we can use the default show() function.
    """
    try:
        get_ipython
        pic.save('temp.svg', *args, **kwds)
        display(SVG(filename='temp.svg'))
    except:
        show(pic, *args, **kwds)

        
r = var('r')     

svg_show(plot(2*r- r^2, r, 0,1) + plot(1, r, 1, 2), ymin=0.5)

[Volker Braun]

The easiest workaround is to switch to vector-graphics output:

%display graphics vector