Releases: Computer-Aided-Validation-Laboratory/pyvale
Releases · Computer-Aided-Validation-Laboratory/pyvale
v2026.2.0
v2026.1.3
v2026.1.1
DIC Improvements:
- Quadratic Shape functions.
- Removed 'Eigen' From Repo.
- Strain moved to separate module. Rename of functions to match between dic and strain modules e.g
dic.calculate_2d()andstrain.calculate_2d(). - Now using
scikit-build-corefor building. - Improved, more informative homemade progress bar.
- General QOL improvements to interactive ROI tool: Seed box now has colour fill.
- Reduced lag when using poly with large number of vertices across large images.
- Updated DIC tests.
- Bugfixes.
Release v2026.1.0
- Small fixes to allow macOS wheels to be built. Now using macos-14, macos-latest and macos-15-intel default github runners.
- Wheels are checked on pull requests from dev -> main.
Release v2025.12.0
What's Changed
- SensorSim Refactor by @ScepticalRabbit in #208
- Breaking changes to the sensor simulation module to streamline sensor array creation with only one obvious way to build things
- Significant improvements to the experiment simulator allowing parallelisation of simulated experiments and IO for saving/loading experiments
- Massive overhaul of the sensor simulation examples in the docs with a minimal set of basic examples and then a much more detailed extended example set. Grouped parts of examples based on workflow steps to make it more intuitive for the user.
- Support for mesh-free sensors
- Support for generic simulation IO to generate SimData objects from txt files or numpy arrays
- Quality of life improvements to the docs splitting out the user guide for different modules with overarching guide to pyvale
- Various bug fixes for the sensor simulation module.
Full Changelog: v2025.8.1...v2025.12.0
v2025.8.1
v2025.7.2
Quality of life and bug fixes:
- allow user to control the number of threads
- ROI tool add pixel coords to the bottom corner of the GUI
- option to output shape function parameters
- nan non-converged DIC values
- issue with initial checks if loading a single deformed image in as .npy
- "dicchecks.py" overly narrow in its acceptance of an integer data type
- allow for save/load of yaml ROI file from code
- Loading yaml seed point seems to flip y axis
- throw an error when the RG seed is outside the ROI
- Output from 'dic_2d' not output into Jupyter Notebook cell output
- removed user input for possible overwrite of results data. Warning provided instead.
- For case where user passes numpy arrays as image input. Assign dummy deformation filenames.
- Added fix for entire ROI usage.