levita_a
b6c96557a0
Fix the object display and also update the interactions so that multi-mode objects work as intended
2026-04-28 14:36:46 +02:00
Abe Levitan
29f0036248
Merge pull request #79 from cdtools-developers/bugfix/saving_loss_metrics
...
Fix a high priority bug where reconstructions will crash upon saving
2026-04-20 16:06:44 +02:00
levita_a
a516f418b9
Fix a high priority bug where reconstructions will crash upon saving, and add test coverage to prevent a repeat
2026-04-20 15:40:55 +02:00
Abe Levitan
65a75b5c7e
Merge pull request #70 from cdtools-developers/newpoisson
...
Switch to a more sensible normalization strategy for loss metrics
2026-04-17 15:12:03 +02:00
levita_a
fcd0b31b2c
Fix a bug with the mask implementation for the normalizers conflicting with SimplePtycho, and update the thresholds to accomodate the new normalization with masks and for intensity_MSE
2026-04-13 15:44:52 +02:00
allevitan
d8fade9f32
Update the amplitude MSE and intensity MSE loss to act equivalently - currently, the intensity MSE loss was by default a mean, but amplitude was by default a sum. Both functions now have a flag, use_sum=False, which is False by default. This changes the default behavior of amplitude_mse. All models, including the tutorial version of simple_ptycho, are updated accordingly and test coverage was added.
2026-04-13 14:34:23 +02:00
levita_a
3c15b19ba7
Find and fix a bug where the normalization code didn't properly include masks
2026-04-10 16:22:58 +02:00
levita_a
3d10f36739
Fix a bug where the normalizers stopped the loss history from being saved as a numpy scalar, and added test coverage
2026-04-05 22:23:23 +02:00
levita_a
388d1669f3
Final checks to ensure all tests pass with the new scaling
2026-04-05 21:52:14 +02:00
levita_a
8ee8bf9237
Fix bug where fancyptycho test was not being run on GPU
2026-04-03 18:59:09 +02:00
levita_a
b6553fd522
Update the example near field ptycho code to use Poisson NLL
2026-04-03 18:48:35 +02:00
allevitan and Claude Sonnet 4.6
3ccdc24d3f
Restructure loss coverage in reconstruction tests
...
- Revert test_Adam_gold_balls to default amplitude_mse
- Add new test_intensity_MSE test (gold balls + AdamReconstructor)
- Update near_field threshold to 17 (poisson_nll scale)
Thresholds for test_intensity_MSE to be tuned after GPU run.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com >
2026-04-03 18:46:27 +02:00
allevitan and Claude Sonnet 4.6
178e9a4e4e
Add loss function coverage to slow reconstruction tests
...
Use poisson_nll in test_near_field_ptycho and intensity_mse in
test_Adam_gold_balls to exercise these loss paths end-to-end.
Thresholds left as-is pending re-running on a GPU machine.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com >
2026-04-03 18:45:29 +02:00
allevitan and Claude Sonnet 4.6
d644782e94
Update poisson_nll test to match sum-based implementation
...
poisson_nll now returns a sum rather than a mean, consistent with the
normalizer pattern. Remove the per-pixel divisions from the numpy
reference calculations accordingly.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com >
2026-04-03 18:44:04 +02:00
allevitan and Claude Sonnet 4.6
fd17ae30ba
Update tutorial to use loss instance attribute pattern
...
Updates tutorial_simple_ptycho.py to match simple_ptycho.py: removes
the def loss() method and instead assigns self.loss and
self.loss_normalizer as instance attributes in __init__.
Updates tutorial.rst accordingly: adds the loss assignment to the
__init__ code block with an explanation, and removes def loss() from
the forward model section.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com >
2026-04-03 18:44:04 +02:00
allevitan and Claude Sonnet 4.6
8c380a34a6
Add intensity_mse loss option to all models
...
Adds 'intensity_mse' as a selectable loss function (alongside the
existing 'amplitude_mse' and 'poisson_nll') to FancyPtycho,
MultislicePtycho, Bragg2DPtycho, Multislice2DPtycho, and RPI.
Models that previously had a hardcoded amplitude_mse assignment now
use the same configurable pattern as FancyPtycho, with the loss
parameter threaded through from_dataset and from_calibration as well.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com >
2026-04-03 18:43:56 +02:00
levita_a
d2c09acd44
Get all tests passing by updating the loss thresholds for the new normalizations, and add the new normalization to all models
2026-04-03 18:38:11 +02:00
allevitan
023ab318ee
Added some better documentation to the loss functions
2026-04-03 18:38:11 +02:00
allevitan
4dbcffd67b
Attempted to update the normalizations for the loss functions in a way which allows for a nice normalization of the Poisson NLL. The slow tests will now likely fail
2026-04-03 18:38:11 +02:00
Abe Levitan
3f028e2fbf
Merge pull request #72 from cdtools-developers/feature/betterplots
...
A bunch of long overdue updates to the plotting system
2026-04-03 18:23:41 +02:00
Abe Levitan
06105d38c7
Merge pull request #76 from cdtools-developers/v0.3.2.dev
...
Start on 0.3.2.dev
2026-04-03 18:22:19 +02:00
levita_a
3b6e73b48c
Start on 0.3.2.dev
2026-04-03 18:20:53 +02:00
allevitan
a87e3415d2
Responding to self-review comments
2026-04-03 17:54:59 +02:00
Abe Levitan
40f829b430
Merge pull request #71 from cdtools-developers/bugfix/intensity_loading
...
Fix issue #68 for loading of datasets with intensities when using OPRP
2026-04-03 16:24:16 +02:00
Abe Levitan
d7d5ed8401
Merge pull request #73 from cdtools-developers/v0.3.1
...
Change version to 0.3.1
2026-04-03 15:43:45 +02:00
Abe Levitan
48a9a7f8ac
Merge pull request #75 from cdtools-developers/readthedocs
...
Update the docs so it doesn't always think it's v0.2.0
2026-04-03 15:29:16 +02:00
allevitan
05bdee3861
Update the docs so it doesn't always think it's v0.2.0
2026-04-03 15:24:50 +02:00
Abe Levitan
e84dc9f76e
Merge pull request #74 from cdtools-developers/readthedocs
...
Add .readthedocs.yaml for ReadTheDocs hosting
2026-04-03 14:32:24 +02:00
allevitan and Claude Sonnet 4.6
7de8448319
Add .readthedocs.yaml for ReadTheDocs hosting
...
Enables versioned documentation on ReadTheDocs. Installs the package
with its docs extras (sphinx, sphinx-rtd-theme, sphinx-argparse) so
autodoc can import cdtools during the build.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com >
2026-04-03 14:28:01 +02:00
allevitan
4030218fa4
Change version to 0.3.1
2026-03-26 15:11:55 +01:00
allevitan
0bd8df498c
Updated the documentation to reflect the changes to the plotting system
2026-03-24 21:39:04 +01:00
allevitan
4172a9a11a
A few small changes to revert unimportant edits and fix linting issues
2026-03-24 19:34:38 +01:00
levita_a
f55b35f0dd
Stop stopping at each plot to show it
2026-03-24 16:56:44 +01:00
allevitan
603486e824
Change the default colormap for exponentiated objects to match that for un-exponentiated objects, and also update test_fancy_ptycho to show all the plots (not just plot level 2)
2026-03-24 16:55:56 +01:00
allevitan
ad2b09f2f0
Update the tests to work better when checking the model plotting, and make sure to cover panel_plot_mode=False
2026-03-24 16:19:26 +01:00
allevitan and Claude Sonnet 4.6
29e9fddd72
Add test coverage for plot_translations, plot_nanomap, and plot_nanomap_with_images
...
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com >
2026-03-24 16:01:45 +01:00
allevitan
25c7b66c7c
Do a final review of all the examples to ensure they work and are well structured
2026-03-24 11:10:23 +01:00
allevitan and Claude Sonnet 4.6
5c5c0be7c6
Reformat plotting function signatures to one-parameter-per-line style
...
All public functions in plotting.py now use the expanded multi-line
signature format with trailing commas for consistency and readability.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com >
2026-03-24 11:00:43 +01:00
levita_a
4201b41573
Made more adjustments to the plotting system to avoid using constained_layout, which was causing hangups during live plotting and didn't look as nice
2026-03-24 09:13:51 +01:00
levita_a
f979388280
Fix an annoying warning coming from double-setting the figsize
2026-03-21 21:42:16 +01:00
allevitan
29656fd78e
Improve the colorized plotting further, and add a colorbar which will be useful for publishing figures now that it's not just a simple hsv lookup. Also fix a bug with nonresponsive windows when all model plots are closed, but the dataset plots are still showing.
2026-03-21 21:27:16 +01:00
allevitan
f04ac9f0ec
Make the colorized plot look nicer, with a more perceptually uniform mapping and a colorbar for accurate reading of phases
2026-03-21 17:35:14 +01:00
allevitan
a4a38530d1
Update the plot_image functions to show sliders
2026-03-21 14:57:16 +01:00
allevitan
f8ff525722
Add example patterns for jupyter notebooks
2026-03-21 08:41:18 +01:00
levita_a
dafb204aa0
Fix a bug where closed windows wouldn't reopen at the original size
2026-03-20 20:18:32 +01:00
allevitan
bef907c032
Added a minimum plotting interval to model.inspect(dataset) to make it easier to not plot so much without manual intervention
2026-03-20 20:01:40 +01:00
levita_a
f35141dcc8
Check that all the models work (all but Multislice2DPtycho, which was already broken, and make necessary tweaks
2026-03-20 19:39:47 +01:00
allevitan
91db5b3c64
Update remaining models to new plot registration system
2026-03-20 19:01:22 +01:00
allevitan
e7f943e254
Made a few more updates to some example scripts, to show the panel plot mode
2026-03-20 17:35:25 +01:00
allevitan and Claude Sonnet 4.6
f4260837bc
Add title option to plot_image and wrappers; misc fixes
...
- Add title parameter to plot_image, plot_real, plot_imag, plot_amplitude,
plot_phase, and plot_colorized
- Various fixes to base.py and fancy_ptycho.py
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com >
2026-03-20 17:20:52 +01:00
levita_a
Abe Levitan