On Wed, 26 May 2010 00:52:24 +0800, Mario Rodriguez <biomates at telefonica.net> wrote:
> Hello,
>
>>
>> ..... and draw3d's mesh function gives incorrect axis values.
> Do you mean how the matrix is projected on the (x,y)-plane? In that
> case, you can reverse de rows of the matrix as in
>
> mesh (reverse(mat), x0, y0, width, height) $
>
>
>
> On the other hand, I'm planning to rename object 'mesh' and call it
> 'elevation_grid', as in VRML. The idea is to call 'mesh' a new more
> general 3d graphic object. Some experimental examples and explanations here:
>
> http://www.telefonica.net/web2/biomates/maxima/gpdraw/3dmodel
>
> I can also change how the matrix is projected on the (x,y)-plane in a
> more convenient way, perhaps following standards in image analysis.
>
> Also, I'm open to suggestions on the naming of the two functions.
>
> Thanks.
> --
> Mario
>
No, not just the projection. When I tried it, there seemed to be an offset in the coordinates.
If I plotted with x0, y0 equal to zero, not data points would actually land on x0, y0, and I think
the width and height of the domain (xmax - x0, ymax - y0) wouldn't equal (width, height). Easy to
work around I guess, and I would have just done that in a pinch, but I was hoping to avoid explicit
discretization just on principle (principle of laziness) :)
Regards,
James