Re: [unclassified] Bug with patch

[Kai Habel]

David Bateman schrieb:
> There is an issue with the patch command.. Firstly it should accept a
> handle as the first argument (consider the attached patch). However the
> more important issue is that it miss treats the case where more that a
> single patch is treated in one call to patch. Consider the code
>
> t = (1/16:1/8:1)'*2*pi;t2 = ((1/16:1/8:1)' + 1/32)*2*pi; x =sin(t); y =
> cos(t); x2 = sin(t2); y2=cos(t2); h = patch([x,x2],[y,y2],'r')
>
> The patches are drawn correctly, however the handle h that is returned
> is a handle to the last patch drawn and not a handle to a multi-element
> patch that includes all of the patches..
>
> Its easy enough to treat the data and get this part right. However,
> looking at the facecolor code I'm not sure what this was originally
> trying to achieve and so I don't see the fix for this immediately.
> Shai/Kai do you a what the facecolor code is trying to do and a way to
> fix it to accept the cdata for multiple facets at once correctly?
>
> D.
>   
When writing my original code I had only single patches in mind. So, I 
have not looked into matlab behavior for different color arguments 
(single float, [rgb], 'color string') using your or a similar example.

But a quick check  shows for a color string and rgb arguments only 
single values are allowed:

h = patch([x,x2],[y,y2],'r')

In that case 'facecolor' contains the rgb value and CData is empty.

For float values you can specify:

h = patch([x,x2],[y,y2],[10 20])

Here 'CData' contains the [10 20] vector, 'FaceColor' is set to 'flat', and 'Clim' is adapted to [10 20] accordingly. In this example those values are new max/min values for CData. 

I hope this help, but I wont be able to provide a patch for this quickly (this 
week).

Kai