W. Trevor King [Wed, 20 Oct 2010 23:48:32 +0000 (19:48 -0400)]
Add autogenerated .be/id-cache to .gitignore.
W. Trevor King [Wed, 20 Oct 2010 23:42:00 +0000 (19:42 -0400)]
Merge branch 'master' of ssh://wking@einstein/~/public_html/code/git/sawsim
W. Trevor King [Wed, 20 Oct 2010 23:35:55 +0000 (19:35 -0400)]
Upgrade .be bug cache to Bugs Everywhere Directory v1.4.
W. Trevor King [Wed, 20 Oct 2010 23:20:14 +0000 (19:20 -0400)]
Disable pysawsim.manager.subproc is multiprocessing is missing.
W. Trevor King [Wed, 20 Oct 2010 23:09:40 +0000 (19:09 -0400)]
Work around namedtuple import for Pythons < 2.6.
W. Trevor King [Wed, 20 Oct 2010 19:57:45 +0000 (15:57 -0400)]
Also cleanup local mpd if the pysawsim nosetests fail.
W. Trevor King [Wed, 20 Oct 2010 19:49:02 +0000 (15:49 -0400)]
Add CC, CFLAGS, and LDFLAGS to the Makefile for easy customization.
W. Trevor King [Wed, 20 Oct 2010 19:23:16 +0000 (15:23 -0400)]
Work around missing MPD with a more elaborate pysawsim nosetest hook.
W. Trevor King [Wed, 20 Oct 2010 17:57:28 +0000 (13:57 -0400)]
Kill off worker-node implementations of master code.
W. Trevor King [Wed, 20 Oct 2010 17:34:48 +0000 (13:34 -0400)]
Add pysawsim.manager.mpi with an mpi4py-based manager.
Test with:
mpdboot -1 -n 1 -f <(hostname)
mpiexec -n 5 nosetests --with-doctest --doctest-tests pysawsim/manager/mpi.py
mpdallexit
I still need to find a way to turn off the doctests when mpi4py is
installed, but the tests are not being run from an `mpiexec`ed
environment.
W. Trevor King [Wed, 20 Oct 2010 11:41:48 +0000 (07:41 -0400)]
Add link to Python wiki's parallel processing page.
W. Trevor King [Wed, 20 Oct 2010 11:30:45 +0000 (07:30 -0400)]
Adjust pysawsim.manager.pbs to skip most doctests if pbs mod is missing.
W. Trevor King [Wed, 20 Oct 2010 10:42:14 +0000 (06:42 -0400)]
Remove execute permission from pysawsim/manager/pbs.py.
W. Trevor King [Wed, 20 Oct 2010 10:38:36 +0000 (06:38 -0400)]
Add pysawsim.manager.subproc using subprocessing.
W. Trevor King [Wed, 20 Oct 2010 10:08:53 +0000 (06:08 -0400)]
Add pysawsim.sawsim.main() and bin/sawsim_hist.py calling it.
Also fix a number of bugs turned up by trial calls to the new
sawsim_hist.py.
W. Trevor King [Wed, 20 Oct 2010 10:05:51 +0000 (06:05 -0400)]
Add Histogram.to_stream() and .headings.
W. Trevor King [Wed, 20 Oct 2010 09:53:17 +0000 (05:53 -0400)]
Fix 'intil_state' -> 'initial_state' typo in pysawsim.sawsim.Event.
W. Trevor King [Wed, 20 Oct 2010 09:51:50 +0000 (05:51 -0400)]
Add Histogram.calculate_bin_edges().
W. Trevor King [Wed, 20 Oct 2010 09:25:50 +0000 (05:25 -0400)]
Reduce ThreadManager default worker_thread to 2.
W. Trevor King [Wed, 20 Oct 2010 08:04:59 +0000 (04:04 -0400)]
Add --sawsim option to velocity_dependant_scan.main().
W. Trevor King [Wed, 20 Oct 2010 08:00:52 +0000 (04:00 -0400)]
Allow Job status logging when .status is an Exception.
W. Trevor King [Wed, 20 Oct 2010 07:57:15 +0000 (03:57 -0400)]
Allow Job.__str__() to work when .id == None.
W. Trevor King [Wed, 20 Oct 2010 07:51:29 +0000 (03:51 -0400)]
Teardown manager after use in velocity_dependant_scan.main().
W. Trevor King [Wed, 20 Oct 2010 07:30:27 +0000 (03:30 -0400)]
Add --clean-cache option to velocity_dependant_scan.main().
W. Trevor King [Wed, 20 Oct 2010 07:21:41 +0000 (03:21 -0400)]
No need for `if __name__ == '__main__'` in velocity_dependant_scan.
bin/vel_dep_scan.py provides access to velocity_dependant_scan.main().
W. Trevor King [Wed, 20 Oct 2010 07:19:01 +0000 (03:19 -0400)]
Rearrange velocity_dependant_scan.main() to group residual-related options together.
W. Trevor King [Wed, 20 Oct 2010 07:16:11 +0000 (03:16 -0400)]
Pass velocity_dependant_scan.main() --manager through to SawsimHistogram.
W. Trevor King [Wed, 20 Oct 2010 07:10:51 +0000 (03:10 -0400)]
Reposition matplotlib.use() call in velocity_dependant_scan.py.
This fixes:
.../matplotlib/__init__.py:833: UserWarning:
This call to matplotlib.use() has no effect because the the backend
has already been chosen; matplotlib.use() must be called *before*
pylab, matplotlib.pyplot, or matplotlib.backends is imported for the
first time.
W. Trevor King [Wed, 20 Oct 2010 07:09:24 +0000 (03:09 -0400)]
Fix sawsim_histogram imports in velocity_dependant_scan.py.
W. Trevor King [Wed, 20 Oct 2010 07:06:05 +0000 (03:06 -0400)]
Add bin/vel_dep_scan.py calling pysawsim.velocity_dependant_scan.main().
W. Trevor King [Wed, 20 Oct 2010 07:05:18 +0000 (03:05 -0400)]
Add --manager option to velocity_dependant_scan.py.
W. Trevor King [Wed, 20 Oct 2010 06:49:17 +0000 (02:49 -0400)]
Add pysawsim.manager.SUBMODS and .IsSubclass and .get_manager.
IsSubclass is from my hooke.util.pluggable module. get_manager() is
based on hooke.util.pluggable.submods() and .construct_odict().
.. _hooke: http://code.google.com/p/hooke/
W. Trevor King [Wed, 20 Oct 2010 01:27:12 +0000 (21:27 -0400)]
Transition more of the pysawsim framework into Python.
W. Trevor King [Wed, 20 Oct 2010 00:55:27 +0000 (20:55 -0400)]
Add pysawsim.sawsim module.
W. Trevor King [Tue, 19 Oct 2010 23:02:19 +0000 (19:02 -0400)]
Move 'qwait' to pysawsim.manager.pbs and add the rest of PBSManager.
W. Trevor King [Tue, 19 Oct 2010 20:43:26 +0000 (16:43 -0400)]
Fix a few deadlock errors in pysawsim.manager.thread.
W. Trevor King [Tue, 19 Oct 2010 14:16:50 +0000 (10:16 -0400)]
Move Histogram class from fit_force_histogram to new pysawsim.histogram module.
W. Trevor King [Tue, 19 Oct 2010 13:04:07 +0000 (09:04 -0400)]
JobManager.async_invoke() should accept Job instances.
It used to attempt to construct them on its own, but with multiple
subclasses (e.g. InvokeJob), that becomes too difficult.
W. Trevor King [Tue, 19 Oct 2010 12:42:54 +0000 (08:42 -0400)]
Add pysawsim.manager and pysawsim.manager.thread for running asynchronous jobs.
Also:
* simplify .gitignore and add *.pyc.
* add log() and __version__ to pysawsim.
* fix exception handling in pysawsim.invoke.invoke().
W. Trevor King [Mon, 18 Oct 2010 21:18:21 +0000 (17:18 -0400)]
Add pysawsim/__init__.py and break invoke out into its own module.
W. Trevor King [Mon, 18 Oct 2010 21:03:51 +0000 (17:03 -0400)]
Add nosetests line to misc/hooks/REAMDE for pre-commit hook.
W. Trevor King [Mon, 18 Oct 2010 20:54:07 +0000 (16:54 -0400)]
Introduce pysawsim in the README and add fit_force_histograms.py & friends.
W. Trevor King [Mon, 18 Oct 2010 20:39:50 +0000 (16:39 -0400)]
Document suggested Git hooks.
`pre-commit-diff-check` is the current stock pre-commit.sample that
`git init` installs.
W. Trevor King [Mon, 18 Oct 2010 19:55:23 +0000 (15:55 -0400)]
"% pop. per s" -> "%% pop. per s" in printf() for k_model_utils.
W. Trevor King [Mon, 18 Oct 2010 19:50:23 +0000 (15:50 -0400)]
Add prelimiary check_* code for each *.h file in sawsim.nw.
W. Trevor King [Mon, 18 Oct 2010 18:00:18 +0000 (14:00 -0400)]
Corrections to sawsim.bib after earlier pybtex run.
Remove extra quotes from around author values
fixed in pybtex revision:
wking@drexel.edu-
20101016180914-98tpuw47ojcl4zsf
Unwrap broken MCarrion-Vazquez macro.
pybtex shouldn't break macros after a '-' character, but it does as
of now.
W. Trevor King [Mon, 18 Oct 2010 13:33:06 +0000 (09:33 -0400)]
Add missing comma between references to tension and rate appendicies.
W. Trevor King [Sun, 17 Oct 2010 21:19:54 +0000 (17:19 -0400)]
Add Tame the BeaST link and von-naming blurbs to README.bibtex.
W. Trevor King [Sat, 16 Oct 2010 16:31:06 +0000 (12:31 -0400)]
Preserve space after \copy macro in sawsim.bib.
W. Trevor King [Sat, 16 Oct 2010 16:28:41 +0000 (12:28 -0400)]
Consolidated ...X macros and abbreviated journal/publisher macros in sawsim.bib.
W. Trevor King [Sat, 16 Oct 2010 16:10:49 +0000 (12:10 -0400)]
Passed sawsim.bib through pybtex and updated keys in sawsim.nw accordingly.
W. Trevor King [Sat, 16 Oct 2010 15:58:52 +0000 (11:58 -0400)]
Corrections to saswim.bib before reformating with pybtex.
W. Trevor King [Sat, 16 Oct 2010 15:58:25 +0000 (11:58 -0400)]
Moved BibTeX comments from src/sawsim.bib to README.bibtex.
W. Trevor King [Thu, 14 Oct 2010 13:58:59 +0000 (09:58 -0400)]
Add examples/Marszalek_Fernandez_1999/titin_intermediate.sh.
W. Trevor King [Thu, 14 Oct 2010 13:48:09 +0000 (09:48 -0400)]
Update copyright/licensing blurbs with newer GPLv3+.
W. Trevor King [Thu, 14 Oct 2010 13:26:42 +0000 (09:26 -0400)]
Added README.git to point Git newbies towards good documentation.
W. Trevor King [Tue, 13 Oct 2009 11:28:59 +0000 (07:28 -0400)]
Added inverse piston comments + whitespace in tension_balance()
W. Trevor King [Fri, 25 Sep 2009 10:52:27 +0000 (06:52 -0400)]
Fixed mispelled 'sec.transtion_rate' reference
W. Trevor King [Fri, 25 Sep 2009 10:43:55 +0000 (06:43 -0400)]
Added cased equation to piston tension model description
W. Trevor King [Fri, 25 Sep 2009 10:30:19 +0000 (06:30 -0400)]
Fixed typo in "Transition rate" section equation
W. Trevor King [Sun, 13 Sep 2009 15:55:50 +0000 (11:55 -0400)]
Bumped to version 0.10.
W. Trevor King [Sun, 13 Sep 2009 15:47:58 +0000 (11:47 -0400)]
Added max dF "-F" option to sawsim.
Also added const_env argument to find_tension(). As I state in a
comment in find_tension(), it should really be a
step_call/new_step/whatever option.
W. Trevor King [Sun, 13 Sep 2009 14:10:18 +0000 (10:10 -0400)]
Added full_chain_stiffness() to handle stiffness with piston tension.
Currently just sets a hardcoded stiffness dx, but that's pretty ugly.
A slightly better solution would be a user-controlled option, but that
is low on my agenda...
W. Trevor King [Sun, 13 Sep 2009 12:13:13 +0000 (08:13 -0400)]
Added piston tension model.
Also typographic cleanup 'assert (' -> 'assert('.
Currently broken: stiffness calculation with the tension model.
W. Trevor King [Fri, 21 Aug 2009 12:57:06 +0000 (08:57 -0400)]
-P option now sets target P_N, not P_1. Bumped to version 0.9.
This cleans up the trouble I had before with
testing/bell_rate/bell_rate.sh 50 1e-6 0.05 1e-3 1e-9 300
where the large number of domains was making P_N much larger than P_1.
This lead to more "ignored double unfoldings", creating a banded
structure in the simulated histogram.
The new '-P sets target P_N' configuration makes it more
straightforward to use -P for it's intended purpose, avoiding "ignored
double unfoldings", without forcing unnecessarily long timesteps once
the population of the limiting state is reduced.
Note that the second purpose of the -P option, avoiding significant
force changes in a single timestep, is either unaffected (for states
with population 1) or improved (for states with population > 1), but
this should perhaps be separately controllable through it's own
"--dF-max" option, or similar...
W. Trevor King [Wed, 19 Aug 2009 16:59:49 +0000 (12:59 -0400)]
Extended bell and const rate tests to multi-domain chains
W. Trevor King [Thu, 13 Aug 2009 18:03:17 +0000 (14:03 -0400)]
The (P_N = NP_1) bug that triggered 0.7 -> 0.8 isn't so bad.
Added notes to that effect. Still, it's better for having thought it
out more clearly ;).
W. Trevor King [Thu, 13 Aug 2009 11:53:04 +0000 (07:53 -0400)]
safe_strto*() now also checks for empty input
W. Trevor King [Thu, 13 Aug 2009 10:09:42 +0000 (06:09 -0400)]
Safer string to number conversion with ato*() -> safe_strto*().
Current hack to avoid including string.h for strlen() in every *.c
file that uses safe_strto*(), which I should fix when I pull them out
into utils.h/c files or similar.
W. Trevor King [Thu, 13 Aug 2009 00:14:01 +0000 (20:14 -0400)]
Removed TODO in favor of BugsEverywhere bugtracking.
See http://bugseverywhere.org for details on BE.
I also removed the following entry from TODO:
maybe histogram without dice rolling for identical domains?
since I decided it was a bad idea. Better to focus the simulation on
what it's good at. Of course, I may change my mind later and
reinstate the issue ;).
W. Trevor King [Thu, 13 Aug 2009 00:06:24 +0000 (20:06 -0400)]
Fixed _major_ bug in multi-domain unfolding calculations & bumped to 0.8.
The relavant lines of code are:
@@ -548,7 +574,7 @@ int domain_transitions(double F, double dt, environment_t *env, int num_domains,
k = accel_k(transition->k, F, env, transition->k_params);
//(*transition->k)(F, env, domain->k_params);
//printf("k = %g,\tdt = %g,\tk dt = %g\n", k, dt, k*dt);
- return happens(k*dt*num_domains); /* N dice rolls for prob. k*dt event */
+ return happens(1-pow((1.0-k*dt), num_domains)); /* N dice rolls for prob. k*dt event */
}
@ [[happens]] is a random decision making function defined in Appendix \ref{sec.utils}.
Yes, I feel very stupid. Sorry.
I also took the oportunity to skim over the explanatory text (up to
the Appendix break), making a few updates and typo corrections as well
as adding more detail to the "Timescales" section discussing
multi-unfolding timesteps.
W. Trevor King [Mon, 3 Aug 2009 20:16:19 +0000 (16:16 -0400)]
inverse_wlc now handles F=HUGE_VAL.
W. Trevor King [Tue, 24 Mar 2009 17:40:02 +0000 (13:40 -0400)]
Added inverse tension functions for speed. Bumped to version 0.7.
W. Trevor King [Mon, 9 Mar 2009 19:06:30 +0000 (15:06 -0400)]
Fixed up the guts to support the new multi-state capabilities.
W. Trevor King [Thu, 5 Mar 2009 18:39:37 +0000 (13:39 -0500)]
Starting on version 0.6, with generalized discrete reactions.
Added domain-chain and reaction-graph figures.
Added timescale section and cleaned up some of the overview text.
W. Trevor King [Wed, 25 Feb 2009 22:27:23 +0000 (17:27 -0500)]
Added stiffness env. parameter and stiffness-corrected Bell model.
Following Walton et al. and my poster. Currently a fairly ugly hack to
get some simulation data for my poster (in 4 days!). I will come
back and clean things up afterwards. Due to the coarseness of the
tension balancer, it's probably a better idea to get stiffnesses
for each of the groups separately and then add the stiffnesses in
series. This removes the balancer completely.
W. Trevor King [Tue, 24 Feb 2009 00:08:38 +0000 (19:08 -0500)]
Increased verbose output from tension_model_utils.
W. Trevor King [Tue, 24 Feb 2009 00:01:23 +0000 (19:01 -0500)]
Corrected some strings for tension_model_utils.
W. Trevor King [Mon, 2 Feb 2009 16:47:09 +0000 (11:47 -0500)]
Fixed a few minor typos.
W. Trevor King [Tue, 11 Nov 2008 20:59:10 +0000 (15:59 -0500)]
Added Fmax and Xmax options to tension_model_utils
Also added Fmax/Xmax/xmax_special/etc current value outputs to
tension_model_utils and k_model_utils.
W. Trevor King [Tue, 9 Sep 2008 17:14:07 +0000 (13:14 -0400)]
Added testing scripts, comparing sawsim with theory
TODO, adjust bell & kramers testing to read param files like
const_rate does.
W. Trevor King [Sun, 7 Sep 2008 19:08:51 +0000 (15:08 -0400)]
Removed debugging printfs from sawsim.c stepping
W. Trevor King [Sat, 6 Sep 2008 01:48:59 +0000 (21:48 -0400)]
Added the option to make discrete steps in x.
W. Trevor King [Tue, 2 Sep 2008 13:21:46 +0000 (09:21 -0400)]
Made directory prerequisites order-only prerequisites.
Avoids excessive rebuilding. See
http://www.kolpackov.net/pipermail/notes/2004-January/000001.html
http://www.gnu.org/software/make/manual/html_node/Prerequisite-Types.html
http://savannah.gnu.org/support/?104349
for details
W. Trevor King [Tue, 2 Sep 2008 12:41:18 +0000 (08:41 -0400)]
Added a better introduction and UTF-8 capability in sawsim.nw.
W. Trevor King [Tue, 2 Sep 2008 11:39:24 +0000 (07:39 -0400)]
Fixed linker error running wlc check in check_sawsim.
Obviously, the wlc check should be in check_tension_model, but that
doesn't exist yet. When it does, it will still need the wlc()
declaration, since wlc() is not exported directly by tension_model.h.
I also added protective #ifndef-#def-#endif protection to the headers.
TODO: Determine which modules depend on which other modules and make
the appropriate includes in the headers themselves. Then the .c files
need only include the headers they use directly.
W. Trevor King [Mon, 1 Sep 2008 03:31:47 +0000 (23:31 -0400)]
Fixed some bugs with `make clean'
I had forgotten to clean up all the sawsim modules :p.
W. Trevor King [Mon, 1 Sep 2008 03:25:36 +0000 (23:25 -0400)]
Major rewrite of the build system/Makefile reorganized.
Almost everything got pushed out into subdirectories of the base
`sawsim/' dir. All the versioned source got sent to `src'.
Processing `sawsim.nw' generates C code, which is placed in `build'
along with the odd bits of code not in the noweb file. Finally, we
compile the code in `build' into executibles placed in `bin' and
documents placed in `doc'. TODO: add appropriate commentary to
`sawsim.nw'.
Standardizing the Makefile also cut down on the module-specific code
needed in the makefile, which should encourage better modulization.
I hadn't been able to compile the checks in a while with the old
system, and now they compile, but they don't pass the checks. Well,
that's what checks are for ;), although in this case I suspect it's
just that I've been ignoring the checks for a while and the code has
moved on, leaving incompatible checks. Fixing the checks is the next
order of business, so we can implement fast, pre-commit sanity-checks.
W. Trevor King [Fri, 29 Aug 2008 23:26:09 +0000 (19:26 -0400)]
Added some missing cleanup targets to the Makefile
W. Trevor King [Fri, 29 Aug 2008 21:04:16 +0000 (17:04 -0400)]
Added a .gitignore file
Noweb produces a lot of unversioned source files. .gitignore keeps
track of them so they don't clutter up your `git status' calls and so
on.
W. Trevor King [Fri, 8 Aug 2008 17:09:16 +0000 (17:09 +0000)]
Fixed a few minor typos
git-svn-id: svn://abax.physics.drexel.edu/sawsim/trunk@15
865a22a6-13cc-4084-8aa6-
876098d8aa20
W. Trevor King [Wed, 6 Aug 2008 10:57:18 +0000 (10:57 +0000)]
Fixed LaTeX typos
git-svn-id: svn://abax.physics.drexel.edu/sawsim/trunk@14
865a22a6-13cc-4084-8aa6-
876098d8aa20
W. Trevor King [Tue, 22 Jul 2008 17:04:21 +0000 (17:04 +0000)]
Fixed some other tension model bugs in tension_model_handler.
git-svn-id: svn://abax.physics.drexel.edu/sawsim/trunk@13
865a22a6-13cc-4084-8aa6-
876098d8aa20
W. Trevor King [Tue, 22 Jul 2008 17:00:40 +0000 (17:00 +0000)]
Passed INIT_MODEL() alterations on to the utility programs.
git-svn-id: svn://abax.physics.drexel.edu/sawsim/trunk@12
865a22a6-13cc-4084-8aa6-
876098d8aa20
W. Trevor King [Mon, 21 Jul 2008 23:53:25 +0000 (23:53 +0000)]
Fixed 'y > yub' error by reducing default tolerances min_d{x,y} in
tension_balance() and x_of_xo(). Now that this is known to be a
problem, I need to come up with a more permanent fix...
git-svn-id: svn://abax.physics.drexel.edu/sawsim/trunk@11
865a22a6-13cc-4084-8aa6-
876098d8aa20
W. Trevor King [Mon, 21 Jul 2008 17:30:57 +0000 (17:30 +0000)]
Fixed group implementation, still need to test it.
git-svn-id: svn://abax.physics.drexel.edu/sawsim/trunk@10
865a22a6-13cc-4084-8aa6-
876098d8aa20
W. Trevor King [Wed, 16 Jul 2008 23:50:01 +0000 (23:50 +0000)]
Added Fmax, xmin, and xmax options to k_model_utils
git-svn-id: svn://abax.physics.drexel.edu/sawsim/trunk@9
865a22a6-13cc-4084-8aa6-
876098d8aa20
W. Trevor King [Wed, 16 Jul 2008 21:32:06 +0000 (21:32 +0000)]
Added noweb dependency information to README
git-svn-id: svn://abax.physics.drexel.edu/sawsim/trunk@8
865a22a6-13cc-4084-8aa6-
876098d8aa20
W. Trevor King [Wed, 16 Jul 2008 19:34:22 +0000 (19:34 +0000)]
Added GSL dependency notes.
git-svn-id: svn://abax.physics.drexel.edu/sawsim/trunk@6
865a22a6-13cc-4084-8aa6-
876098d8aa20
W. Trevor King [Wed, 16 Jul 2008 19:26:56 +0000 (19:26 +0000)]
Added interpolating lookup table files. They are from my interp_table package
(a GNU libavl spinoff), so they aren't included in sawsim.nw.
git-svn-id: svn://abax.physics.drexel.edu/sawsim/trunk@5
865a22a6-13cc-4084-8aa6-
876098d8aa20