William Stein wrote: > On Thu, Oct 22, 2009 at 10:13 AM, Jason Grout > <[email protected]> wrote: >> Bill Page wrote: >>> The optional labeled axes in jmol 3-d graphics is a nice addition but >>> take a look at, for example: >>> >>> line3d([(0,0,0), (-3,4,-5),(3,4,-5),(3,-4,5)]) >>> >>> In the 4.1.2 notebook. Right-click the image and turn on Style/Axes. >>> Notice how the diagonal line does not appear to start at where you >>> would expect the origin to be. The origin appears to be displaced a >>> little along the Y-axes. Both the X and Y axes appear to extend beyond >>> the box. Not too serious but is this a known problem? >> >> Yes. The axes you turn on are centered in the picture, but do not >> correspond to data coordinates. >> >> The underlying reason has to do with some fiddling that happens with >> rescaling things so that to jmol, everything appears to be in a box >> between [-1..1] (or something like that). The axes appear at the jmol > > The motivation for using rescaling things and *not* using jmol > coordinates, is that things go totally to hell if we use jmol > coordinates. E.g., in math it makes perfect sense to say draw > something where the x and y axis span a range of say 10^(-16), or to > draw a sphere of radius 10^(-10), etc. If you do that with jmol > coordinates, disaster occurs, since everything is floating point (??). > So what happens now is that Sage maps its own scene to something > like [-1,1] and jmol then renders it all nicely. I was the one > who wrote the code so that this mapping is used, since I got very > frustrated with jmol seriously corrupted images. > > You can understand why jmol wouldn't care about uses funny scales > (like an x-axis range of 10^(-16)), because jmol is a program for > visualizing molecules after all! >
Bob recently suggested (on the jmol list) creating a special mathematics isosurface object. Specifically, he said: "...I suggest we start talking about a new object type. Maybe a "plot3d" that is an extended isosurface. We already have that for molecular orbitals, LCAO cartoons, and pmesh. The way it works is that the extended object has more capabilities -- in this case, to map more than just contours onto an isosurface. Like you say, a totally variable mesh, which would just be a set of lines of variable width, color, and translucency. Possibly annotations. In addition, possibly, automatic scaling. Right now you have to do some transforms that are oblivious to Jmol. I'd like the oblivious part to be in the use interface and let Jmol take care of anisotropic scaling." I posted about this here: http://groups.google.com/group/sage-devel/browse_frm/thread/c737c188e6342a8d# So, if anyone has any comments on what is needed in a "mathematical plot" object for jmol, please post! Jason -- Jason Grout --~--~---------~--~----~------------~-------~--~----~ To post to this group, send an email to [email protected] To unsubscribe from this group, send an email to [email protected] For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
