Skip to content

Conversation

@RichardScottOZ
Copy link
Contributor

Just a bit easier to read

Also doubt you are going to do a demo video? ;)

RichardScottOZ and others added 30 commits May 26, 2023 20:30
Co-authored-by: RichardScottOZ <72196131+RichardScottOZ@users.noreply.github.com>
Co-authored-by: RichardScottOZ <72196131+RichardScottOZ@users.noreply.github.com>
Co-authored-by: RichardScottOZ <72196131+RichardScottOZ@users.noreply.github.com>
Co-authored-by: RichardScottOZ <72196131+RichardScottOZ@users.noreply.github.com>
- Create postprocessing.py with PostProcessor class
- Integrate DBSCAN clustering and line/curve fitting
- Add predict_and_postprocess method to ModelPredictor
- Create POSTPROCESSING_GUIDE.md with detailed documentation
- Add postprocessing_example.py demonstrating full pipeline
- Support Linear, Curve, and BestCurve fitting methods
- Include visualization and statistics capabilities

Co-authored-by: RichardScottOZ <72196131+RichardScottOZ@users.noreply.github.com>
- Create PIPELINE_COVERAGE.md documenting all original features
- Add bridge.py for seamless original→modern integration
- DatasetAdapter for using original DATASET with modern models
- FilterAdapter for accessing rotation filters
- LegacyTrainer combining original data with modern architectures
- Add integration_example.py demonstrating all integration patterns
- Update README with integration section
- Document backward compatibility and feature coverage

Co-authored-by: RichardScottOZ <72196131+RichardScottOZ@users.noreply.github.com>
…ly-pipeline

Modernize LineamentLearning: TensorFlow 2.x, multi-architecture support, CLI, post-processing, integration bridge, and comprehensive documentation
Co-authored-by: RichardScottOZ <72196131+RichardScottOZ@users.noreply.github.com>
Co-authored-by: RichardScottOZ <72196131+RichardScottOZ@users.noreply.github.com>
Co-authored-by: RichardScottOZ <72196131+RichardScottOZ@users.noreply.github.com>
…ix type hints

Co-authored-by: RichardScottOZ <72196131+RichardScottOZ@users.noreply.github.com>
…g-rotation

Implement data loading and rotation augmentation integration
- Created MAT_TO_PYDATA_GUIDE.md with detailed conversion documentation
- Added mat_converter.py module with conversion utilities for NumPy, HDF5, and Zarr
- Added mat_conversion_examples.py with practical examples
- Updated DATASET.py to support loading from .npz and .h5 files
- Updated data_generator.py to support multiple file formats
- Updated README.md with data conversion information
- Updated examples/README.md with conversion quick start

Co-authored-by: RichardScottOZ <72196131+RichardScottOZ@users.noreply.github.com>
- Fixed convert() method to properly filter kwargs for each format
- Added comprehensive testing with synthetic dataset
- Verified conversion, validation, and loading from all formats
- All tests pass: .mat, .npz, and .h5 formats work correctly

Co-authored-by: RichardScottOZ <72196131+RichardScottOZ@users.noreply.github.com>
- Clarified Parquet is documentation-only, not fully implemented
- Fixed unused prefix parameter in recursive HDF5 loading
- Added Zarr format validation support
- Cleaned up recursive function definitions

Co-authored-by: RichardScottOZ <72196131+RichardScottOZ@users.noreply.github.com>
…or-mat-to-pydata

Add .mat to PyData conversion utilities and multi-format data loading
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant