Log plots of points are blank when given small values

[scma...@gmail.com]

This is mentioned in http://trac.sagemath.org/ticket/13430#comment:1, but it is later stated that this is fixed by http://trac.sagemath.org/ticket/13528. However, it appears that this latter ticket was never applied, as the problem still exists in version 5.12. Since the ticket listing the problem was closed nine months ago, I'm not sure what the proper way to bring up this issue is (whether to re-open that ticket, start a new one, comment on the ticket that supposedly fixes the issue, etc), so I'm asking about it here.

For reference, here is the issue:

sage: list_plot_semilogy([10^-i for i in range(3)])

works, but

sage: list_plot_semilogy([10^-i for i in range(4)])

does not (it produces a blank plot).

I encountered this issue using different data that included very small y values (order 10^-12). Since you can't include points on the x-axis for semilogy plots, it is possible that points that are too small are being interpreted as 0. Consider the following:

sage: list_plot_semilogy([10^-i for i in range(3)]+[0])

produces a blank plot, but

sage: list_plot_semilogy([10^-i for i in range(3)]+[0], plotjoined=True)

produces a line connecting the first three points (and apparently ignoring the fourth point). However,

sage: list_plot_semilogy([10^-i for i in range(6)], plotjoined=True)

produces a line connecting all six points.

[kcrisman]

On Thursday, November 14, 2013 12:09:39 PM UTC-5, scma...@gmail.com wrote:

This is mentioned in http://trac.sagemath.org/ticket/13430#comment:1, but it is later stated that this is fixed by http://trac.sagemath.org/ticket/13528. However, it appears that this latter ticket was never applied, as the problem still exists in version 5.12.

It shouldn't have been closed, and isn't.  It is positive review, pending review of a ticket it depends on - http://trac.sagemath.org/ticket/13422 - which I haven't had/made time to review properly because of some confusion I probably caused with a weird question.  If you are able to review this (see the Sage developer guide) then we can get both of them in, which would be marvelous!

- kcrisman