Releases: weigertlab/spotiflow
Releases · weigertlab/spotiflow
0.6.4
- Improve inference tiling to alleviate border effects.
- Allow
min_distanceto be a floating point number. - Adjust
min_distanceautomatically if model grid is greater than 1. - Added support for Python 3.14
- CI refactor.
Full Changelog: 0.6.2...0.6.4
0.6.2
- Fix lazy normalization in prediction CLI
- Faster sum mode for heatmap generation
- Fix wrong data loading argumens in notebook examples (#48)
- New artwork
Full Changelog: 0.6.1...0.6.2
0.6.1
- Disable MPS for 3D models
Full Changelog: 0.6.0...0.6.1
0.6.0
- Added new
fluo_livemodel for single molecule tracking - Override
sigmawhen fine-tuning trained models - Flexibilize CLI, more Zarr prediction options
- Move model hosting to GitHub (#43).
- Few minor bug fixing
Full Changelog: 0.5.8...0.6.0
0.5.8
What's Changed
- Added support for OME-Zarr prediction in CLI
- Add
--channelsargument to training and prediction CLI for channel selection by @qin-yu in #39 - Example notebook fix
- Fix 3D Gaussian fit
- Bump Windows version for CI
Full Changelog: 0.5.7...0.5.8
Bugfix: 3D multi-channel training
New Contributors
Full Changelog: 0.5.6...0.5.7
Bugfix: multi-channel training
- Fixed bug causing multi-channel training not to work
- Fixed bug where training crashed if an explicit
SpotiflowTrainingConfiginstance was passed as thetrain_paramsargument to theSpotiflow.fit()method. - Fixed tests.
Full Changelog: 0.5.5...0.5.6
0.5.5
What's Changed
- Remove verbosity argument in scheduler by @anwai98 in #30
- better augmentation api;data loading by @maweigert in #29
- fixed model 3d training argument
- CLI defaults to point priority sampling for training
New Contributors
Full Changelog: 0.5.4...0.5.5
0.5.4
- Add support for
numpy>=2.0(#28) - Add support for Python
3.13
Full Changelog: 0.5.3...0.5.4
0.5.3
- Fixed #26, #27
- Default default to more likely sampling of patches containing spots (at least 80% of them)
- Added 2D+t sample data
Full Changelog: 0.5.2...0.5.3