Behavior of subplot

Michael D. Godfrey godfrey at isl.stanford.edu
Sat Feb 14 23:04:32 CST 2009


> After thinking about if further, the changes due to the presence of 
> x/ylabel and the title are not relevant to the subplot function. Once 
> the proper listeners are in place the axes position/outerposition will 
> be modified as the axes labels/title are added.
>
> Thus, the only change needed is to respect the behavior of subplot 
> when the default axes units are changed.
>
> I've made some appropriate changes and attached a new changeset.
I just tried this latest version.  I have attached before and after 
PostScript of one of my
uses of subplot:

1. tt2.ps   is before changeset
    tt4.ps   is after.

My comments are:

1. The Bounding Box problem is fixed.
2. The title and xlabel issues are a bit worse. And, the
     difference between what is shown on the (X11) screen and
     what goes into the PS is greater.  The overlap of the xlabel
     and titles is a little worse on the screen, but the overlap completely
     in the PostScript.
3.  Making the horizontal spacing of the plots closer, as you have done,
     is probably a good idea (and may be "officially" correct) but it has
     had the effect that my subplot titles overlap.  I can fix this, of 
course.

So, I would say that if you can improve the xlabel/title overlap and make
the screen layout and PostScript layout match better  (PostScript more like
the screen), this will be good to submit as far as I know.

Also, I like the tic marks in the latest better!!

-------------- next part --------------
A non-text attachment was scrubbed...
Name: tt2.ps
Type: application/postscript
Size: 65849 bytes
Desc: not available
Url : https://www-old.cae.wisc.edu/pipermail/bug-octave/attachments/20090214/d4d80a23/attachment-0002.ps 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: tt4.ps
Type: application/postscript
Size: 65910 bytes
Desc: not available
Url : https://www-old.cae.wisc.edu/pipermail/bug-octave/attachments/20090214/d4d80a23/attachment-0003.ps 


More information about the Bug-octave mailing list