v1.3.2

[JudithBernett]

PR Checklist

• This comment contains a description of changes (with reason)
• If you've fixed a bug or added code that should be tested, add tests!
• Documentation in docs is updated

Description of changes

• Preprint is out! Linking it in the README
• Bump to Python 3.13 (now supporting 3.11, 3.12, 3.13)
• New versions, separate pre-commit action
• Memes on the documentation page!
• CurveCurator: name the new data columns like the constants CELL_LINE_IDENTIFIER (=cell_line_name), DRUG_IDENTIFIER (=pubchem_id) instead of "cell_line_id", "drug_id"
• response from_csv: look for measure, cell line identifier, and drug identifier column; cast drug identifier to str; same for to_dataframe.
• feature from_csv: forgot metainfo, some fixes
• load_custom, evaluate_file: forgot dataset_name
• final_model_on_full data: if True, a final/production model is saved in the results directory. If hyperparameter_tuning is true, the final model is produced according to the hyperparameter tuning procedure, which was evaluated in the nested cross-validation. Accordingly, all models now have save() and load() functions.
• Transformation on the features like scaling, variance feature selection, … are now implemented as sklearn transformer so that they can be fitted in train and applied for predicting. Because of that, cell line/drug features are now copied in experiment.py
• Added SingleDrugProteomicsRandomForest, ProteomicsElasticNet
• Made Naive Models nicer/less redundant
• Added no_hyperparameter_tuning flag
• Forcing NaiveMeanEffectsPredictor to be always run