summary |
shortlog | log |
commit |
commitdiff |
tree
first ⋅ prev ⋅ next
W. Trevor King [Mon, 27 May 2013 16:30:04 +0000 (12:30 -0400)]
README: Remove duplicate pyafm target
W. Trevor King [Mon, 27 May 2013 10:52:47 +0000 (06:52 -0400)]
README: Update to discuss unfold.py script and workflow
The old instructions were from back when I ran unfolding scans from
the IPython prompt.
W. Trevor King [Mon, 27 May 2013 10:18:33 +0000 (06:18 -0400)]
config: No need to import FFT_tools.window_hann
W. Trevor King [Fri, 24 May 2013 21:31:56 +0000 (17:31 -0400)]
Convert to load_from_config() handling
Rather than juggling independent configurations, consolidate them
under ScanConfig in the style of pyafm.storage. This makes it easy to
load a full scanning config from ~/.config/unfold-default.yaml, which
is a plain-text file that can easily be tweaked by hand.
To generate an initial unfold-default.yaml file, try something like:
>>> import unfold_protein.storage as storage
>>> config = storage.get_default_config()
>>> storage.save_scan_config(config=config)'
which uses defaults from unfold_protein.config as well as the default
AFM configuration from pyafm.storage.load_config().
W. Trevor King [Fri, 24 May 2013 21:26:08 +0000 (17:26 -0400)]
config: Use native Python types for default unfolding velocities
This avoids the need for funkyness when saving the configuration with
h5config's YAML storage backend.
W. Trevor King [Fri, 24 May 2013 19:27:04 +0000 (15:27 -0400)]
Add packaging boilerplate
W. Trevor King [Fri, 24 May 2013 19:22:27 +0000 (15:22 -0400)]
.be: Upgrade bugs-everywhere repository from v1.0 to v1.5
W. Trevor King [Sun, 12 May 2013 20:36:54 +0000 (16:36 -0400)]
plot-unfold.py: Autoscale tight (to avoid wasting space)
W. Trevor King [Sat, 11 May 2013 15:14:17 +0000 (11:14 -0400)]
unfold.py: Remove config overrides for unfolding distance, etc.
These snuck in during
87f12e4 (Update unfold_protein to use the new
pyafm.storage._load_afm, 2012-05-11). Obviously, I should be using
h5config to load a user-written YAML file instead of hard-coding these
values in the unfolding script.
W. Trevor King [Wed, 13 Mar 2013 14:03:41 +0000 (10:03 -0400)]
plot-unfold.py: Add a simple h5 -> png converter
This makes it easy to filter successful pulls by eye using your
favorite image viewer.
W. Trevor King [Thu, 7 Feb 2013 06:13:42 +0000 (01:13 -0500)]
scan: Add stepper_tweaks (--no-stepper-tweaks) to UnfoldScanner.run()
This is useful if you're not around to monitor a long run. If the
laser is disrupted (e.g. via a bubble in the fluid cell), the
photodiode voltage can flatlign, even if you aren't off the surface.
A confused UnfoldScanner with *think* it's off the surface, and charge
blindly in, crushing the cantilever into itsy bits ;).
Laser disruption detection is a great reason to measure the total
photodiode voltage. Unfortunately, that signal is not carried on the
cable between the MultiMode and the NanoScope. I tried jumping it out
of the MultiMode by hand, but after a few weeks like that, the board
died. I'm not sure it was related, but I think it's safer to just
disable stepper tweaks when you're not around to monitor the signal.
W. Trevor King [Thu, 7 Feb 2013 06:07:34 +0000 (01:07 -0500)]
scan: Approach the surface with stepper_approach() before looping
This makes the initial approach much faster if you are starting far
from the surface (e.g. after a calibcant calibration).
W. Trevor King [Thu, 7 Feb 2013 06:05:55 +0000 (01:05 -0500)]
scan: Log the loop index
So the user can gauge the remaining experiment time.
W. Trevor King [Thu, 7 Feb 2013 06:03:00 +0000 (01:03 -0500)]
unfolder: Save the timestamp, x-position, and temp under environment/
Keep the HDF5 root clean by moving these parameters under the
environment/ group. Earlier versions of unfold-protein did not record
the x-position.
W. Trevor King [Thu, 7 Feb 2013 06:01:25 +0000 (01:01 -0500)]
Standardize matplotlib rendering on figure.canvas.draw() and figure.show().
In interactive mode, skip the no-op pyplot.show(). In non-interactive
mode, the show() calls will block until the window is closed.
W. Trevor King [Fri, 11 May 2012 18:22:38 +0000 (14:22 -0400)]
Ran update-copyright.py.
W. Trevor King [Fri, 11 May 2012 18:21:40 +0000 (14:21 -0400)]
Add .update-copyright.conf.
Use my update-copyright project to manage copyright blurbs.
http://pypi.python.org/pypi/update-copyright/
W. Trevor King [Wed, 18 Apr 2012 17:24:54 +0000 (13:24 -0400)]
Change my email address from drexel.edu to tremily.us.
W. Trevor King [Fri, 11 May 2012 17:42:47 +0000 (13:42 -0400)]
Update unfold_protein to use the new pyafm.storage._load_afm.
This is a messy commit to get my Git repo in sync with the version I'm
currently using in the lab, but most of the changes have to do with
the nested-config overhaul starting with pypiezo's:
commit
de497a3734372f5fa7c92d6ff7bdb6b2e327c345
Author: W. Trevor King <wking@drexel.edu>
Date: Thu Mar 15 10:13:20 2012 -0400
Optional config-based-setup for PiezoAxis, OutputChannel, and InputChannel.
See the module docstring for details on why this is useful.
W. Trevor King [Mon, 26 Mar 2012 18:53:32 +0000 (14:53 -0400)]
Remove TemperatureConfig from unfold_protein.config.UnfoldCycleConfig.
Replaced by pyafm.config.AFMConfig['temperature'].
W. Trevor King [Mon, 26 Mar 2012 18:50:51 +0000 (14:50 -0400)]
Remove unfold_protein.temperature (replaced by pyafm.temperature).
W. Trevor King [Mon, 26 Mar 2012 18:32:02 +0000 (14:32 -0400)]
Remove unfold_protein.afm now that there is a configurable default AFM in pyafm.
Now you can use pyafm.storage.load_afm() and instead of hardcoding your
configuration in the unfold_protein.afm module.
W. Trevor King [Wed, 25 Jan 2012 15:33:43 +0000 (10:33 -0500)]
Update unfold.py config to show temp in Kelvin.
By default, `calibcant.config.TemperatureConfig` assumes you'll give
temperatures in Celsius, but in
`unfold_python.temperature.Temperature.get_temperature`, we convert
the backend temperature from Celsius to Kelvin.
W. Trevor King [Tue, 24 Jan 2012 13:08:42 +0000 (08:08 -0500)]
Assorted plots and bug-fixes in unfolder.py.
* Updated `ExceptionTooFar` plot and `Unfolder._plot` to use `FIGURE`
and `_pyplot`.
* Fix configs used in `Unfolder._unfold`'s bit/meter conversions.
* Remove unrecognized keyword `dtype` from the `.astype` call.
* Reshape `out` so it is 2D (as required by `Piezo.ramp`).
* Don't force unfold log values to be integers.
* Fix filename generation in `Unfolder._save`.
* Removed trailing whitespace.
W. Trevor King [Tue, 24 Jan 2012 13:02:08 +0000 (08:02 -0500)]
Make ExceptionTooFar check more robust with .max().
W. Trevor King [Tue, 24 Jan 2012 12:59:04 +0000 (07:59 -0500)]
Stylistic cleanups in UnfoldScanner.move_far_from_surface.
W. Trevor King [Tue, 24 Jan 2012 12:57:50 +0000 (07:57 -0500)]
Increase 'far' in ApproachConfig to 30 microns.
W. Trevor King [Mon, 23 Jan 2012 19:29:05 +0000 (14:29 -0500)]
Remove buggy self argument from Temperature self() call.
W. Trevor King [Tue, 17 Jan 2012 22:53:03 +0000 (17:53 -0500)]
Merge irrelevant "Update to data_logger v0.4." by dropping the changes.
The new h5config framework no longer uses data_logger to locate or
save data.
W. Trevor King [Tue, 17 Jan 2012 22:49:26 +0000 (17:49 -0500)]
Update to use new h5config, pycomedi, etc.
W. Trevor King [Tue, 17 Jan 2012 22:04:34 +0000 (17:04 -0500)]
Add .gitignore.
W. Trevor King [Tue, 9 Nov 2010 20:14:58 +0000 (15:14 -0500)]
Update to data_logger v0.4.
W. Trevor King [Mon, 6 Jul 2009 20:23:26 +0000 (16:23 -0400)]
Avoid banded non-contact region by reducing default stepsize.
See bug
18dd6d0a-3c45-4aba-80da-
fa32ffaaf5db for example banded data.
(plot deflection without lines to see why I call it "banded"). Reducing
the stepsize reduces the jolt on the cantilever, and reducing the
points-per-step accordingly means we don't have to change the
card-clock rate. It might be worth adding "target-clock-rate" option,
and letting the stepsize be determined automatically (as an
alternative to setting the stepsize explicitly) but we'll
put that off for now...
W. Trevor King [Fri, 26 Jun 2009 18:55:20 +0000 (14:55 -0400)]
Added README file.
Also a bug/thoughts on packaging trouble (blocked by temperature).
W. Trevor King [Fri, 26 Jun 2009 18:32:20 +0000 (14:32 -0400)]
Updated to work with newly bundled pizeo package.
W. Trevor King [Wed, 3 Jun 2009 19:22:41 +0000 (15:22 -0400)]
Added EFILE debugging output to ExceptionTooClose case.
W. Trevor King [Wed, 3 Jun 2009 19:19:40 +0000 (15:19 -0400)]
Began versioning.