Clinica

Changed

** Core: **

* Add ClinicaDL installation with pipx.
* Improve logging.
* Add method argument to the interpret command to choose between the new Grad-CAM method and the gradient method.
* Change `extract` command to `prepare-data`.
* Change output of `get-labels`, `split` and `kfold` commands to one TSV per split instead of one per label.
* Change `tsvtool` command to `tsvtools`.
* Change `tsvtools getlabels` command to `tsvtools get-labels` and remove the progression column in the TSV output.
* Add new commands: `tsvtools get-progression`, `tsvtools get-metadata`, `tsvtools prepare-experiment` (split + kfold), and `tsvtools adapt`.
* Update data CI.
* Add a new model for quality check

Fixed

- Fix `quality-check t1-linear`

Changed

** Core: **

- Dependencies versions were updated in order to support Python equal or
greater than 3.8.
- Security updates for some dependencies.
- Include the `--save_tensor` and `--save_nifty` options into the `predict`
sub-command (before `save_tensor` was also subcommand`).
- Update some paper references in the documentation.
- Add a description for the available models.
- Add the flag `--model_layers` to the `clinicadl train list_models` subcommand
to visualize the model layers.

Fixed

- Fix bug when using the `clinicadl train list_models` command.

Changed

** Core: **

- Add new VAE models for training.
- Add multiprocessing for the `extract` task.
- Add tests to validate `resume` and `quality-check` functionalities.
- Update minimal version for first level dependencies: clinica, scikit-learn,
scikit-image.
- Add a new Makefile to the root of the repository in order to facilitate
sysadmin frequent tasks. The CI script was also modified.

** Command line:**

- Update of the train command line: now all actions related to training
(resume, list models, tasks…) are sub-commands of the command `clinicadl
train`. Task is now a sub-command, and not an argument.
- Add command to list the available models for the training task and the
details of each model.
- Add an option to the command `save-tensor` to save the reconstruction
outputs as nifti files.
- Add command `clinicadl train from_json` to train again the same experiment
from a JSON file generated by ClinicaDL.

Fixed

- Allow the user to choose the optimizer and the metric independently of the
performed task.
- Logger does not display output according to the `--verbose` option during
the training tasks. This behavior is fixed.
- Behavior of `resume` was changed.
- Reader to request BIDS/CAPS datasets changed its interface in the recent
version of Clinica (>=0.5.4). Code using these functions was adapted.
- Exceptions raised are now more specific and detailed.
- Documentation was updated.

Changed

** Core: **

Switch to Poetry to manage production and development requirements.
Update requirements.
Improve metrics for multi-class settings.
Remove NaN values in input data.
Allow to choose the name of the JSON file produced by `extract`.
Improve logs.

** Command line:**
Add `selection_metrics` option in `clinicadl train` (already available in TOML file).
Print help if no argument is given.

Fixed

Fix label code generation
Remove all occurrences of a group in a MAPS when choosing the `overwriting` option.
Fix `resume` when the training.tsv file is empty.

ClinicaDL 1.0.3 (release bugfix)

Fixed

- Fix import module for VAE architectures.

Changed

**Core:**

- Clinica version dependency.
- Update documentation, with proofreading.
- Change the test for `extract`.

Fixed

- Use the GPU option on the command line interface.
- Fix the train task when invoking the CLI.
- Replace the `predict` arguments passed through the CLI.
- Fix issue when using multiple cohorts.

Changed

**Functionalities:**

- `extract` improve this functionality with more expressive and easy to understand flags.
- Change the test for `extract`.