On 2010-07-09 15:12:02 +0200, Gael Varoquaux said: > I hate having people stuck. If you have specific suggestions for better > documentation, I very gladly take them.
OK, here are some suggestions and questions to the documentation of the surf function: > s is the elevation matrix, a 2D array. Fine. > x and y can be 1D or 2D arrays (such as returned by numpy.ogrid or > numpy.mgrid), but the points should be located on an orthogonal grid > (possibly non-uniform). In other words, all the points sharing a same > index in the s array need to have the same x or y value. For > arbitrary-shaped position arrays (non-orthogonal grids), see the mesh > function. Quite confusing, why does it not matter if x or y are 1D or 2D arrays? What happens in either case? Can I mix it? (x as 1D , y as 2D)? I don't understand 'sharing a same index in the s array', respectively the 'need to have'. Do you mean: All permutations between the given x and y array points are made, and only when they exacly represent a coordinate of the s array, that value is taken and being drawn? If you could explain this to me, I am quite willing to work on doc improvements. Obviously I need to understand it well for that to happen. ;) But maybe we should communicate this outside the list now? BR, Michael > > Gael > > ----- Original message ----- > > On 2010-07-09 09:51:09 +0200, Gael Varoquaux said: > > > > > On Fri, Jul 09, 2010 at 09:42:03AM +0200, K.-Michael Aye wrote: > > > > This 2nd way I have tried many times, as I am working completely in > > > > numpy arrays, but I always get flat object, like this: > > > > http://dl.dropbox.com/u/139035/snapshot.png > > > > > > Ah, you need to use: > > > > > > mlab.surf(img, warp_scale='auto') > > > > > > Right now, the z scaling (also called 'warp_scale') is calculated in > > > the same units than (x, y) pixels. > > > > Ahhh, NOW I understand that sentence! ;) > > What happened is, that my pixel values (=luminance) were around 0.1, > > which created a surface plot of the height of around 0.1 pixels, i.e. : > > flat. > > > > Thanks for making it click! But maybe one could improve the description > > of the z-scaling a bit better, improve clarity and maybe to include > > this little number example with 0.1 as pixel values? I have the feeling > > it would help all newbies to understand how to do the scaling. > > >> BR, >> > Michael >> > >> > > >> > > HTH, >> > > >> > > Gaël >> > > >> > > >> ------------------------------------------------------------------------------ > >> >> > This SF.net email is sponsored by Sprint >> > > What will you do first with EVO, the first 4G phone? >> > > Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first >> > >> > >> > >> > >> > >> ------------------------------------------------------------------------------ > >> >> This SF.net email is sponsored by Sprint >> > What will you do first with EVO, the first 4G phone? >> > Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first >> > _______________________________________________ >> > MayaVi-users mailing list >> > MayaVi-users@lists.sourceforge.net >> > https://lists.sourceforge.net/lists/listinfo/mayavi-users >> >> >> ------------------------------------------------------------------------------ >> This SF.net email is sponsored by Sprint >> What will you do first with EVO, the first 4G phone? >> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first >> _______________________________________________ >> MayaVi-users mailing list >> MayaVi-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/mayavi-users ------------------------------------------------------------------------------ This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first _______________________________________________ MayaVi-users mailing list MayaVi-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/mayavi-users
