I would just like to give a +1 on the importance of fixing this. I am often irritated by my explicit bounds being expanded in hard to predict ways. I am somewhat confused by the way the 2D graphics use and alter matplotlib, so I didn't fix it myself either, but I could at least review a patch.
The first thing a lot of people want to do with something like Sage is plot something, and if it doesn't look good they might not try it again. I have literally seen that happen. The 3D graphics have been a huge boost but 2D is what people try first. -M. Hampton On Apr 10, 6:15 am, "Joel B. Mohler" <[EMAIL PROTECTED]> wrote: > On Wednesday 09 April 2008 07:37:56 pm gerhard wrote: > > > * The actual bounds used for the axes > > can yield very surprising results. > > Should the range bounds be honored if > > the user sets them explicitely? > > Since we had a thread about this a bit ago and I was allegedly going to write > a patch, I'll share my results of the patch. I had a patch which (in the > case I was interested in) eliminated 3 places which nudge the margins about > for aesthetic reasons. I thought it might be a manageable job to turn that > into a more generic patch which honors margins. > > However, I was thoroughly scared off when I saw what would need to be done to > honor the bounds setting for all cases (frame/no frame, etc). There is a lot > of "magic constants" in the code (0.1, 0.015, 0.002, etc) which evidently > were labored over for days. This is a very worthwhile task, but it sure > looks like it needs a more massive overhaul than I wanted to engage in. > > In general, I think the right algorithm is something like this: > 1) Honor the bounds settings > 2) Check if there are tick marks, axis labels, tick labels outside the > viewable area > 3) Enlarge the viewable area using font metrics to include these labels. > > I know nothing about matplotlib so parts 2&3 are not possible for me in any > amount of time I wanted to commit to this. However, parts 2&3 are necessary > for the frames=True option. I don't know if it is actually difficult or not. > I totally reject the "aesthetic" padding idea though so it really is (IMO) > just a matter of finding font metrics from whatever code is making text. > > -- > Joel --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
