Asymptote plotting

[kcrisman]

Still need improvement on plotting of asymptote of e.g. tan or 1/x, especially when detect_poles != True. Ideally one has an output similar to

sage: plot(tan,-20,20).show(ymin=-5, ymax=5) 

but just straight out of plot.

[mabshoff]

Hmm, I am wondering if #3907 fixed this issue?

Cheers,

Michael

[kcrisman]

I doubt it, because e.g. the tangent example above usually won't get 'inf' or '-inf' as something in ydata; it just has weird things happen near asymptotes because of the adaptive plotting, and also will occasionally get random points that are really far from y=0. I'll try it when 3.1.2 comes out, though.

It's actually kind of fun to evaluate the above command again and again, seeing what you get each time... If you try a few times you can get a ydata point that is 10000 or even more, which requires only that plot "guesses" a number within 10^-5 of pi/2 for the xdata...

[kcrisman]

Just FYI, #6035 partially resolves this.

However,

sage: plot(tan,-20,20, detect_poles='show')
sage: plot(tan,-20,20, detect_poles=True)

shows that 'show' still keeps the "missing" points (which include not just the asymptote but also the high slopes on either side) and that True doesn't create a uniform height as one might desire in the Description.

Description changed to indicate this successful partial resolution, however!

[kcrisman]

See also #8341.