May 2003 - Chimera-dev - RBVI Mailing Lists

Xform
by Thomas Goddard 13 May '03

13 May '03

Hi Steve, The most common way to set an Xform is with the rotate() and translate() methods. You can also use the following Xform method static Xform xform(Real r00, Real r01, Real r02, Real t03, Real r10, Real r11, Real r12, Real t13, Real r20, Real r21, Real r22, Real t23, bool orthogonalize = false); You need to call it as chimera.Xform_xform() from Python. This underscore notation is what we are using for static methods. Perhaps we will provide chimera.Xform.xform() when Python 2.3 with class method support comes out. The getRotation() and getTranslation() Xform methods let you read out the transform. Unfortunately there is no method to directly read the Xform matrix elements because our Python wrapper program can't handle returning arrays. I will ask Greg Couch if there is anything that can be done to remedy this. I see the chimera/include directory I gave you had symbollic links to alot of header files. I've made that a new http://www.cgl.ucsf.edu/home/goddard/chiminc.tar.gz that includes all the files. The Xform class is defined in include/otf/Geom3d.h, but the real file that determines what Xform methods are available in Python is called pyinterface. I included it in the new chiminc.tar.gz as include/_chimera/pyinterface. Look at its Xform definition to find out what the Python methods are. Tom > Date: Mon, 12 May 2003 23:15:32 -0500 (CDT) > From: Steve Ludtke <stevel(a)blake.3dem.bioch.bcm.tmc.edu> > Reply-To: sludtke(a)bcm.tmc.edu > To: Thomas Goddard <goddard(a)cgl.ucsf.edu> > > Well, things are getting a lot better. The headers and tips you provided > are a quantum leap forward. The otf headers weren't in the tarball you > sent, but I found otf0.2 on your website and downloaded that. My problem > now is access to the geometry. The Xform class is a nice abstraction of > the transformation matrix, but I haven't found a good way to directly > modify or extract the matrix elements. I'd like to generate a sequence of > Xform matrices myself, but aside from converting them into > rotation/translation sequences, I don't see a way to set the values. Even > more important I don't see a good way to read out the matrix values. There > is a print method, but it's rather annoying to have to convert the matrix > to ascii only to convert it back into a 'real' matrix again. Is there a > better way I'm missing ? > > Thx for the continued assistance :^) > > -- > ---------------------------------------------------------------------------- > Steven Ludtke, PhD | Baylor College of Medicine > sludtke(a)bcm.tmc.edu | Co-Director > stevel(a)alumni.caltech.edu | National Center For Macromolecular Imaging > V: (713)798-6989 | Dept of Biochemistry and Mol. Biol. > instant messenger: sludtke42 | * Those who do ARE * > http://ncmi.bcm.tmc.edu/~stevel | The converse also holds

1 0

Re: Chimera at NCMI
by Thomas Goddard 12 May '03

12 May '03

Hi Steve The Chimera programming documentation is abysmal. We usually use the C++ header files which have equivalent Python interfaces. Unfortunately you don't have those in the distribution. I've packaged up the Chimera include directory. You can get them at http://www.cgl.ucsf.edu/home/goddard/chiminc.tar.gz We keep them in /usr/local/chimera/include. You'll want to mostly look at include/_chimera/*.h. Also include/otf/Geom.h is helpful. A) how to rotate/translate the camera >>> v = chimera.viewer >>> c = v.camera >>> print c.center (5.9539999961853027, -2.186500072479248, 10.296500205993652) >>> c.center = (x, y, z) # to translate camera >>> # Camera always points in -z direction. There is no way to rotate it. >>> v.scaleFactor = 1.5 # to zoom camera B) how to rotate/translate individual objects C) how to find out the current rotations/translations >>> om = chimera.openModels >>> mlist = om.list() >>> m = mlist[0] >>> xf = chimera.Xform() >>> axis = chimera.Vector(1, 0, 0) >>> angle = 90 # degrees >>> xf.rotate(axis, angle) >>> print m.openState.xform # 3x3 rotation matrix # last column is translation 0.982695 0.121524 0.139793 -1.07064 0.0250348 0.660639 -0.750287 6.83425 -0.183531 0.740803 0.646164 6.35578 >>> m.openState.globalXform(xf) # Another method # xf.premultiply(m.openState.xform) # changes xf # m.openState.xform = xf # To rotate relative to data axes use # m.openState.localXform(xf) # or # xf.multiply(m.openState.xform) # changes xf # m.openState.xform = xf D) how to save the current view to a file (haven't worked on this very hard yet, but while I'm asking) >>> v = chimera.viewer >>> im = v.pilImage() # ImagingCore object >>> from PIL import Image >>> image = Image.Image()._makeself(im) >>> format = 'PNG' # Or 'BMP', 'EPS','GIF', 'JPEG', 'PDF', 'PPM', 'TIFF' >>> path = '/home/goddard/hoohoo.png' >>> image.save(path, format = format) ---------------------- > Date: Sun, 11 May 2003 23:21:48 -0500 (CDT) > From: Steve Ludtke <stevel(a)blake.3dem.bioch.bcm.tmc.edu> > Reply-To: sludtke(a)bcm.tmc.edu > > Hi again. I decided to take another stab at doing some scripted animations > with chimera, but got stuck. I've looked through the available docs pretty > thoroughly, and I've spent a lot of time interrogating a live session, but > there are still some serious holes, largely because a lot of the code is > in C and I don't have the source on that side. Most of the docs seem to > focus on manipulating the actual data rather than the viewing framework. > I know how to get access to the list of open models and to the current > lensviewer. However, I still can't figure out: > > A) how to rotate/translate the camera > B) how to rotate/translate individual objects > C) how to find out the current rotations/translations > D) how to save the current view to a file (haven't worked on this very > hard yet, but while I'm asking) > > If you could send me any sort of code snippet indicating how to do these > things, I'd be grateful. I'm sure they are pretty much one-liners, but I > can't figure out what the 1 line is... > > thx again! >

1 0

compression
by Eric Pettersen 07 May '03

07 May '03

I've been thinking a little more about using compressed files. How about this: Current OpenSave functions would continue to work much as they do now, except that appropriate .gz files would also be listed in the browser, and possibly returned in the list of paths. Instead of opening returned paths with the built-in open, you would instead use a function in OpenSave (osOpen?) that would check for possible compression and return an appropriate file-like object. For contexts where an uncompressed file path is needed (e.g. handing off to PDBio [for now]), there would be another function in OpenSave that takes a path and returns a corresponding path to a uncompressed file [a no-op if the file was already uncompressed]. Any temporary uncompressed files would automatically be deleted on APPQUIT. Opinions? --Eric

1 1

alt-loc funcs
by Eric Pettersen 01 May '03

01 May '03

There are now two new attributes in Atom to assist in working with alternate locations: Atom.allLocations returns a list of atoms sharing the atom's location, from highest occupancy to lowest. In the most common case (no alternate locations) it is a list with one element, the current atom. Atom.primaryBonds is similar to Atom.bonds, but holds at most one bond to a location. If this atom has an alternate location indicator, then primaryBonds will hold bonds to atoms with the same indicator, or no indicator. If this atom has no indicator, primaryBonds will hold bonds to the highest-occupancy atoms of bonded locations. --Eric

1 0

today's chimera development meeting
by Tom Ferrin 01 May '03

01 May '03

Conrad's now got what Amelia has. He said he was even too sick to type on his keyboard! Eric is hereby designated to run today's Chimera Developer's meeting. The agenda topics are listed on Conrad's white board. I'll try to sit in for at least part of the meeting. --tom

1 0

Chimera-dev May 2003