v0.3.1

Release v0.3.1

Major Features

Minor Improvements

• Upgraded Python version to 3.7.3.
• Upgraded TensorFlow version to 1.13.1.

Bug Fixes and Other Changes

• Fixed typo in image mapper.

v0.3.0

Release v0.3.0

Major Features

• Added FACT, H.E.S.S.-I, H.E.S.S.-II, and MAGIC cameras to ImageMapper.
• Added bilinear interpolation, bicubic interpolation, nearest neighbor interpolation, rebinning, image shifting, and axial addressing image mapping methods in ImageMapper.
• Added support for running models using data of multiple telescope types.
• Added use_peak_times data loading option to load peak arrival times from data files.

Minor Improvements

• Added auto_configuration.py script to automatically change the paths in benchmark configuration files.
• Added argument in run_multiple_configurations.py to resume from a particular run.
• Added summarize_results.py to summarize the results of a set of runs.
• Rationalized the metadata variables returned by DataLoader.
• Added test_image_mapper.ipynb for testing the image mapping methods of ImageMapper.
• Changed telescope names for compatibility with ctapipe camera names.
• Refactored ImageMapper to implement all image mapping methods as matrix operations, so that more expensive calculations are performed only during initialization.

Bug Fixes and Other Changes

• Fixed DivisionByZeroError in apply_cuts during HDF5DataLoader initialization.
• Renamed internal variables in and generally cleaned up DataLoader.
• Refactored DataLoader._load_metadata() into smaller functions for clarity and efficiency.
• Fixed incorrect logging of array examples by class.
• Changed model loading to use included CTLearn models by default.
• Added contributing guidelines.
• Made run_model.py append the CTLearn version number to config files.
• Updated TensorFlow version to v1.12.
• Added benchmark configuration files for CTLearn v0.3.0.
• Removed deprecated models.

CTLearn v0.2.0

Release v0.2.0

Major Features

• User-defined TensorFlow classification models with custom configuration parameters can now be imported, in addition to the Single Telescope, CNN-RNN, and Variable Input Network models provided with CTLearn.
• Image mapping added for all CTA telescope types as well as VERITAS.
• Data loading, data processing, and image mapping have been refactored into separate classes with methods to load HDF5 files, preprocess generic IACT data, and map telescope data to square images. Each class is defined in a separate module.
• Configuration now uses YAML instead of INI format, allowing lists and dictionaries to be configured cleanly.

Minor Improvements

• Benchmark configuration files and results have been produced using Single Telescope and CNN-RNN models for all CTA telescope types.
• Package installation now uses a conda environment file to resolve dependencies, providing clean and light installation and removal.
• Training and prediction now handled as two modes within run_model.
• Updated run_multiple_configurations.py to allow configuration parameter combinations to be grouped together.
• Image mapping is now configurable with options for padding and hexagonal conversion algorithm.
• Prediction output is now NumPy-compatible and includes the run number, event number, and telescope ID (if applicable) of each example.

Bug Fixes and Other Changes

• Renamed project to CTLearn from CTALearn.
• Added BSD 3-Clause license.
• TensorFlow version updated to v1.9.0.
• Clarified telescope sorting options.
• Fixed errors in and otherwise updated supplementary scripts.
• Moved unsupported models to models/deprecated and will be removed in the next release.
• Removed plot_gpu_util.py.
• Added workaround to handle overflow error in tel_id parameter of ImageExtractor HDF5 file format.
• Removed dependency on TensorFlow-Slim.

OSX Support

This release updates the requirements to provide support for OSX.

Hands-on Session

This release includes three main improvements. First, the TensorFlow version has been updated to the most recent version v1.7. Second, the CNN-RNN model has been updated and improved and can be used for classification. Third, a script has been added for prediction so that trained models can now be applied to test data. In addition, several supplementary scripts are provided to plot ROC curves using the predicted classification values.

Initial Release

First pre-release version of ctalearn. This release is intended to provide a basis for code validation on different machines.

Notes

• As with all v0.x releases, this is a development release. Future versions may include substantial revisions and breaking changes.
• This release uses Tensorflow v1.4.1. The Tensorflow version will be upgraded in a future release.
• The correctness and performance of the available models have not yet been fully validated. Use at your own risk.
• If you would like to test the code on your machine, please contact the repository authors for a tarball containing a configuration file and Tensorflow checkpoints for a benchmark run.