- Chimera-dev - RBVI Mailing Lists

Re: [chimera-dev] [Chimera-users] Non-gui chimera in a shell?
by Eric Pettersen 09 May '06

09 May '06

What Tom says below is all true, but as he intimated in his previous mail if you are willing/able to use Chimera as your Python interpreter (typically as "chimera --nogui script.py") then things should work. There are some caveats. Chimera ignores the PYTHONPATH environment variable in order to avoid using libraries from incompatible Python installations, therefore to have Chimera find locally installed packages that you want to use you need to set the CHIMERAPATH variable, which Chimera uses in a fashion similar to PYTHONPATH. Another thing is that you can't pass arguments to your script in the conventional fashion since Chimera itself processes all arguments. If you need to pass values to your script, you need to use environment variables as detailed in the FAQ here: http:// www.cgl.ucsf.edu/chimera/docs/ProgrammersGuide/faq.html#q3 --Eric Eric Pettersen UCSF Computer Graphics Lab pett(a)cgl.ucsf.edu http://www.cgl.ucsf.edu On May 8, 2006, at 9:27 PM, Thomas Goddard wrote: > Hi Jim, > > Importing Chimera from your Python script to use its > functionality is > something we have put very little effort into supporting. I have done > this in an nuclear magnetic resonance assignment program called Sparky > that is based on Python to run Chimera as a structure viewer. > Below is > the code that starts Chimera from the Sparky Python interpretter. > > There are many problems. > > You need to set an environment variable to tell the operating system > where the Chimera shared libraries are before you start Python. See > example below. > > You need to use the Python included with Chimera because Chimera > uses a modified Python. Ick. The modification allows Chimera objects > like Molecules to be given new attributes. There is also some other > issue on the Mac where a stock Python can't find the init_chimera() > method in shared library _chimera.so. > > If you start the Chimera graphical user interface then you will need > to run the Tk event-loop to get Chimera to update the graphics. But > then Chimera will not return to your script that started it. So my > example below does not start the event-loop. > > If you start Chimera in nogui mode, it quits as soon as it processes > any arguments which would probably prevent your script from calling > into > Chimera after that as Chimera has cleaned up. > > When I used the script below and then opened a PDB file on my Mac OS > 10.4.6 system the screen went dim and a message told me to hold the > power button down to do a hard reboot! Probably a graphics driver > issue. > > As you can see we really haven't worked much on this type of > Chimera use. > > Tom > > ---- > > Example starting Chimera on a Mac from Python: > > $ export CHIMERA="/Users/goddard/Desktop/Chimera 2199.app/Contents/ > Resources" > $ export DYLD_LIBRARY_PATH=$CHIMERA/lib > $ "/Users/goddard/Desktop/Chimera 2199.app/Contents/Resources/bin/ > python2.4" > Python 2.4.2 (#1, Jan 30 2006, 11:15:23) > [GCC 4.0.1 (Apple Computer, Inc. build 5247)] on darwin > Type "help", "copyright", "credits" or "license" for more information. >>>> import start_chimera >>>> start_chimera.start_chimera() >>>> >>>> # Now you could "import chimera" and do some stuff. >>>> > > --- > > The following is the start_sparky.py file used in the above example: > > # > ---------------------------------------------------------------------- > ------- > # > start_count = 0 > def start_chimera(): > > setup_chimera() > > import chimeraInit > try: > chimeraInit.init(argv = [], nogui = False, eventloop = False, > exitonquit = False) > except SystemExit, value: > # Chimera raises SystemExit sometimes when it can't start. > # Have to catch it here or Sparky will exit. > import traceback > traceback.print_exc() > raise EnvironmentError, 'Chimera failed to start' > except: > global start_count > if start_count == 1: > warning = ''' > Chimera can only be started once per Sparky session. > This is a limitation in all versions of Chimera released so far > (through March 2004). > > ''' > import sys > sys.stdout.write(warning) > raise > > start_count += 1 > > # > ---------------------------------------------------------------------- > ------- > # Do initial Chimera setup that only needs to be done once even if > Chimera > # is started and quit several times. > # > chimera_setup = False > def setup_chimera(): > > global chimera_setup > if chimera_setup: > return > chimera_setup = True > > import os > if not os.environ.has_key('CHIMERA'): > raise EnvironmentError, 'CHIMERA environment variable not set.' > > chimera_path = os.environ['CHIMERA'] > import os.path > if not os.path.exists(chimera_path): > raise EnvironmentError, 'CHIMERA environment variable gives non- > existent directory ' + chimera_path > > chimera_python = os.path.join(chimera_path, 'share') > chimera_lib = os.path.join(chimera_path, 'lib') > import sys > if sys.platform == 'win32': > chimera_site_packages = os.path.join(chimera_path, 'bin', 'Lib', > 'site-packages') > else: > chimera_site_packages = os.path.join(chimera_path, 'lib', > 'python2.4', > 'site-packages') > > sys.path.insert(0, chimera_site_packages) > sys.path.insert(0, chimera_python) > sys.path.insert(0, chimera_lib) > > _______________________________________________ > Chimera-users mailing list > Chimera-users(a)cgl.ucsf.edu > http://www.cgl.ucsf.edu/mailman/listinfo/chimera-users

1 0

Benchmark results
by Daniil Zaitsev 24 Apr '06

24 Apr '06

Graphics ATI Mobility Radeon X1300 Processor Intel Core Duo T2300 Clock (MHz) 1700 OS Type Windows XP Media Edition Surface 394 Mesh 486 Contour 141 Solid 256 Recolor 88 Notes Laptop Test Date 4/06 Chimera Version 1.2199 Machine Dell Inspiron E1505 (laptop) Graphics Memory (Mb) 128 Main Memory (Mb) 512 OS Version Windows XP Media Edition Driver Version Host name

2 1

for Eric Petterson
by Bill Gleason 17 Apr '06

17 Apr '06

Hi Eric, Could you please reply to this directly so that I have your email address? Couldn't find the "direct mail address." We are having some weird problems with reopening the .py session files. Can't really put a finger on it, but every once in a while a .py file gets written and it cannot be opened by Chimera. Not all the time, mostly it is ok, but it can be really annoying when it happens. We are running (mostly) Windows although also some Linux boxes. I can't really remember this happening on Linux. Thanks for any comments. Bill -- William B. Gleason, University of Minnesota Associate Professor, Department of Laboratory Medicine & Pathology Fellow, University of Minnesota Supercomputer Institute Graduate Faculty: Chemistry, Medicinal Chemistry, and Biomedical Engineering (612) 626-2455 (office) (612) 626-5115 (lab) www.umn.edu/~bgleason

1 0

Re: [chimera-dev] [Chimera-users] Rotation about an axis
by Thomas Goddard 11 Apr '06

11 Apr '06

Hi Jordan, I see one mistake in your code, related to a confusing aspect of Chimera transform objects. mvo = apply(chimera.Vector, origin) vo = ref_model.openState.xform.apply(mvo) Here you use Xform.apply(Vector). Xform is a rotation and a translation but when applied to a Vector you get only the rotation applied. That is what you want if you are reorienting an axis, but not if you are repositioning the origin. The Xform.apply(Point) method applies both the rotation and translation. So you want: mvo = apply(chimera.Point, origin) vo = ref_model.openState.xform.apply(mvo) In fact, that apply() call is old style Python and I would use: mvo = chimera.Point(*origin) vo = ref_model.openState.xform.apply(mvo) The * in front of the argument means each element of the sequence origin is passed as a separate argument. This change messes up the translate() lines further down: xf.translate(vo) xf.rotate(va, angle_deg) xf.translate(-vo) The Xform.translate(Vector) requires a Vector argument not a Point. So I'd change that to use Xform.translate(x,y,z) as follows xf.translate(vo.x, vo.y, vo.z) xf.rotate(va, angle_deg) xf.translate(-vo.x, -vo.y, -vo.z) An important detail to note is that the transform xf when applied to a point first translates by -vo, then rotates, the translates by +vo. That is the order is backwards from the above function calls that created the transform. (In the ContrainedMove code there is an incorrect comment about translate(u) actually translating by -u -- I didn't realize what was going on when I wrote that code.) If I understand your code correctly you are specifying a line (origin and axis) to rotate about in the coordinate system of ref_model, then expressing that line in Chimera global coordinates, and then rotating the second model about that line. There are so many pits to fall into with the Xform / Vector / Point Chimera code. And there is not documentation for it, other than reading the source code. This is especially excruciating because it is hard to even find the source code. I looked at Chimera source code file chimera/libs/_chimera/pyinterface to see the Python interface to the C++ Xform, Vector, and Point objects. And then when I needed to actually look at the Xform C++ code I found that in otf/lib/geom3d/Xform.cpp which is in the Chimera foreign source code distribution (200 Mb) in a tar ball: chimera/foreign/otf/otf.tar.gz This source code is available from the Chimera web site. Could we make it any more difficult for you? It's a shameful mess. Feel free to ask further questions. There is a Chimera developers mailing list chimera-dev(a)cgl.ucsf.edu that is more appropriate for coding questions. Tom

1 0

new chimera.fileInfo.register interface
by Greg Couch 08 Mar '06

08 Mar '06

There are now two additional keyword arguments to chimera.fileInfo.register: category and dangerous. The category should be set to one of the built-in categories or one you invent. The categories are meant to help organize all of the file types chimera can open. A set of default categories has been created, chimera.FileInfo.X where X is DYNAMICS for a molecular trajectory (amber), GENERIC3D for generic 3D objects (vrml, bild), SCRIPT for command scripts (python, midas), SEQUENCE for molecular sequences, STRUCTURE for molecule structure (pdb, mol2), SURFACE for molecular surfaces (grasp), and VOLUME for volume data. The dangerous flag defaults to True for scripts and false otherwise. You can override that default by giving the keyword. So far, the categories have only been added to the default file types in the chimera module and to the Midas module. Please update your uses of chimera.fileInfo.register in the extensions you maintain. - Greg

2 1

chimera.widgets.ModelOptionMenu aggressiveness
by Charlie Moad 06 Mar '06

06 Mar '06

I am using chimera.widgets.ModelOptionMenu in my plugin as suggested, but when I attempt to save a screen image its as if my plugin/the structure is being unmapped then mapped again. The tiles are rendered before my plugin reloads the interface. Is there a way I can programmatically destroy my plugin's instance so I can at least take a screenshot without anything changing? I poked around chimera.extension.manager without success. Thanks, Charlie

2 1

1.2199 and BILD
by Charlie Moad 28 Feb '06

28 Feb '06

First of all, sorry for posting all those extension questions on the user list. I just discovered the dev list. Could I get the headers needed for 1.2199 for linux and possibly OS10.4? I am wanting to make use of the BILD format and I have a few questions. If I generate a lot of polygons how efficient will it be, or is there some way to specify poly-strips/triangle-fans. Pertaining to PDB's as well, is there a way to open a file from a stream/string? I despise temp files. It would be nice to use a StringIO object to write my contents to, or do something like "chimera.openModels.openBuffer(myBuffer, type='PDB')", where myBuffer is anything with a read() method or maybe just a string. Thanks, - Charlie

4 4

publish collaboratory idl files?
by David E. Konerding 03 Feb '06

03 Feb '06

Hi Chimera developers, I was wondering if it would be possible to publish the Collaboratory IDL files? I would like to make a remote control application for chimera, but don't want to use the puppet extension. It seems like a nice way to do this would be to use the Collaboratory interface. Thanks, Dave

2 3

Fwd: Re: Tkinter_Idle askyesno issue
by Wolf Cochrane 13 Jan '06

13 Jan '06

My local Chimera python is Python2.3, and it works fine when I run the file with it. It is only when I run the file through the Chimera Idle version that it doesn't work. Wolf ----- Forwarded message from Conrad Huang <conrad(a)cgl.ucsf.edu> ----- Date: Thu, 12 Jan 2006 09:14:04 -0800 From: Conrad Huang <conrad(a)cgl.ucsf.edu> Reply-To: Conrad Huang <conrad(a)cgl.ucsf.edu> Subject: Re: [chimera-dev] Tkinter_Idle askyesno issue To: Wolf Cochrane <s24449424(a)tuks.co.za> Can you please try the same thing with Chimera's version of Python? It's in $CHIMERA/bin/python2.4, where $CHIMERA is where you installed Chimera. Thanks, Conrad On Jan 12, 2006, at 12:01 AM, Wolf Cochrane wrote: > Hi, > > I still have the same bug. The same file works fine if I run it > with my local > version of Python(on Suse 9.2), but when run through the Chimera > Idle version it > always returns "False". > > > > #------------------------------------------------------ > import Tkinter > import tkMessageBox > > print tkMessageBox.askyesno("Dialog","Click Yes or No") > > #------------------------------------------------------ > > > cheers > Wolf Cochrane > > > > > > > > > > > > > > tried this on Chimera 1.2184 on a Windows XP system and the code > returns True when you click Yes and False when you click No. > > Conrad > > On Jan 11, 2006, at 8:45 AM, Wolf Cochrane wrote: > >> Hi >> >> Is it just me of does this bit of code always print "False" when > >> run through >> Idle (Chimera v1.2065). If so, how to I get around it without >> having to build my >> own dialogs. >> >> >> >> import Tkinter >> import tkMessageBox >> >> #root = Tk() >> #root.geometry("500x510") >> #root.title("Yes/No") >> >> #root.mainloop() >> >> >> print tkMessageBox.askyesno("Linker range","Click Yes if a > range of >> Linkers will >> be used at this Permutation Site") >> >> >> >> >> Thanks >> Wolf >> >> _______________________________________________ >> Chimera-dev mailing list >> Chimera-dev(a)cgl.ucsf.edu >> http://www.cgl.ucsf.edu/mailman/listinfo/chimera-dev > _______________________________________________ > Chimera-dev mailing list > Chimera-dev(a)cgl.ucsf.edu > http://www.cgl.ucsf.edu/mailman/listinfo/chimera-dev ----- End forwarded message -----

1 0

Re: [chimera-dev] Tkinter_Idle askyesno issue
by Wolf Cochrane 12 Jan '06

12 Jan '06

Hi, I still have the same bug. The same file works fine if I run it with my local version of Python(on Suse 9.2), but when run through the Chimera Idle version it always returns "False". #------------------------------------------------------ import Tkinter import tkMessageBox print tkMessageBox.askyesno("Dialog","Click Yes or No") #------------------------------------------------------ cheers Wolf Cochrane tried this on Chimera 1.2184 on a Windows XP system and the code returns True when you click Yes and False when you click No. Conrad On Jan 11, 2006, at 8:45 AM, Wolf Cochrane wrote: > Hi > > Is it just me of does this bit of code always print "False" when > run through > Idle (Chimera v1.2065). If so, how to I get around it without > having to build my > own dialogs. > > > > import Tkinter > import tkMessageBox > > #root = Tk() > #root.geometry("500x510") > #root.title("Yes/No") > > #root.mainloop() > > > print tkMessageBox.askyesno("Linker range","Click Yes if a range of > Linkers will > be used at this Permutation Site") > > > > > Thanks > Wolf > > _______________________________________________ > Chimera-dev mailing list > Chimera-dev(a)cgl.ucsf.edu > http://www.cgl.ucsf.edu/mailman/listinfo/chimera-dev

2 1