Releases: AdvancedPhotonSource/ptychodus
Releases · AdvancedPhotonSource/ptychodus
v1.3.2
- migrate from gladier and fair-research-login to directly use globus-sdk
- fix probe orthogonalization bug that was introduced in previous release
- move data visualization methods to API
- switch to colorcet for perceptually uniform color maps
- improve forward model for synthetic diffraction pattern generation
- improve API documentation & test coverage
v1.3.1: Expand Python API
- Added position/probe/object generator methods to API
- Cleaned up Globus module
- Updated LYNX automation plugin
- Fix hidden pty-chi widgets
- Added detector pixel geometry to assembled diffraction data
- Added user prompt to sync parameters on settings save
- Bugfix: refresh diffraction info text after allocating patterns array
- Added procedural object builders: Gaussian random field and fractal/simplex noise
v1.3: Remote training & inference
- update packaging
- improve API documentation & update docs
- add diffraction and product file I/O to API
- reduce default logging verbosity
- update trainable reconstructor interface
- grow workflow API to expose more actions, including registering products and transforming probe positions
- update automation modules to use task manager
- update globus modules to use task manager; add remote inference & training workflows
- update processing modules to support background training tasks
- update processing view to simplify GUI
- consolidate controller timers to simplify synchronization tasks
- add diffraction pattern simulator
- add dead leaves procedural object builder
- illumination dialog can display different quantities
- improve numerics in XMCD analyzer
- improve HDF5 file filtering in LYNX automation plugin
- bugfix: diffraction metadata is now applied before assembling array
- random object builder supports blur
- add fallback data source to Polar position file reader
- update PtychoPINN integration and enahance SLAC file readers
v1.2: Improve background task handling (#116)
- introduce “task manager” to improve synchronization of data loading and reconstruction tasks; implement blocking option for diffraction data loading and reconstruction tasks
- several breaking API changes: renamed scan -> probe positions based on user feedback, add option to disable diffraction pattern processing
- implement reconstruction progress monitoring
- fix bad pixels handling and add documentation for good/bad pixel masks
- simplify behavior of crop center editing widgets
- closing diffraction dataset clears visualization
- rework stxm analysis dialog to stxm object builder
- improve object fourier analysis dialog and promote from developer mode
- extract globus integration from workflow module
- rename ptychoshelves readers -> fold_slice readers; fix loss value loading
- change default hdf5 file writer compression method to lzf
- fix fresnel zone plate probe builder for rectangular probes
- display current object and probe pixel sizes in GUI
- remove tike integration
- add convert-to-ptychodus script for translating prepared datasets to ptychodus file format
- update ptychouds-bdp script and batch mode actions to use standard directory layout
- standardize argument naming across scripts
- add convert-to-ptychodus and ptychodus-bdp to project scripts
v1.1.2
v1.1.1
- prototype object Fourier analysis dialog
- change background of editable fields in item views
- make diffraction view more responsive by caching diffraction pattern summaries
- update PtychoPINN integration to match recent API changes
- update file readers: APS Polar, APS ISN, LCLS XPP, MAX-IV NanoMax
- better debugging information for troubleshooting HDF5 compression plugins
- fix loss values I/O in HDF5 & NPZ product reader/writer
v1.1.0: v1.1 (#111)
- fix issue with numpy 2.3 vectorize
- changed default file format from NPZ to HDF5
- add readthedocs documentation: ptychodus.readthedocs.io
- update pty-chi integration to support v1.2 api
- add readers for multiple NSLS-II data formats
- read new metadata in ISN diffraction file reader
- read new metadata in LYNX diffraction file reader
- display depth of field and detector numerical aperture in product editor
- renamed "patterns" modules to "diffraction"
- python api changes
- product api
- rename costs to losses and use LossValue to record epoch with value
- add exposure_time_s to metadata
- reconstruction api
- add logging to ReconstructorLibrary interface
- update train output to return training loss and validation loss with epochs
- workflow api
- added method to get prepared reconstruct input
- added method to train reconstructor
- product api
- prototype object fourier analysis dialog
v1.0.2: Update to pty-chi 1.1 (#110)
- Update to pty-chi v1.1 api
- Enable loading bad pixels masks from file
- Add sphinx documentation
- Add file readers for APS Polar (4-ID), Ptycho-SAXS (12-ID-E), and ISN (19-ID) instruments
- Update file reader names to be more consistent
- Add support for transposing diffraction patterns
- Bundle xdrlib so MDA file readers work with python 3.13+
v1.0.1
v1.0: pty-chi (#103)
- Add integrations for pty-chi and ptychopinn
- Updated template globus configuration file for ALCF Polaris
- New file readers: cSAXS, NSLS-II, SLAC, CSSI, ISN
- Improve diffraction pattern loading: enable binning, padding, and update wizard
- Add diffraction pattern counts meter to GUI
- Implement minimal support for spatially varying probes
- Added ability to rescale model probe power when specified photon count > 0
- Add ability to estimate probe photon count from diffraction data
- Packaging: Update packaging to support wheels and create Dockerfile
- Add position/probe/object file reader adapters for product file readers
- Update analysis dialogs
- Add macro to attempt reconstructions using all standard scan transforms
- Support auto-loading data products from file
- Reconstructions: execute in different thread, add status widget so GUI doesn’t lock up during reconstructions
- Update memory usage display to use LCD widget