jeandut

Change default value for timeout

3

Hi ! Currently the default timeout is 100 seconds. For inattentive users the abrupt stop of the computation is surprising and hard to debug. Maybe like 10 minutes (or 600...

Unexpected stop because of check_convergence in stopping_criterion

2

When running benchmarks with multiple solvers and the params max_runs set, some solvers will end up producing a much lower number of iterates than max_runs because `check_convergence` makes them exit...

refactor: modifying _update_from_checkpoint signature

1

## Related issue No related issue ## Summary ## Notes This PR introduces a change in the signature of the algorithm's `update_from_checkpoint` function in order to ease its use in...

This information is already available from formulaic directly. When `design` is your design matrix, there is ``` design.model_spec.variable_terms ``` and ``` design.model_spec.term_variables ``` to map from variables to formulae terms...

**Is your feature request related to a problem? Please describe.** As of now comparison between pydeseq2 and deseq2 necessitates to run deseq2 in R/r2py then save results in csvs in...

Caching preprocessed features in Kits19

1

As the preprocessing from `run_nnUnet_preprocessing.py` is super RAM intensive and long we need to cache saved features and allow restarts to use previously computed features.

Especially for datasets that require special evaluation functions such as Fed-LIDC and Fed-KITS there is a lack on documentation on which function to use and what it monitors.

Doc enhancement: explain more clearly datasets and associated hyperparameters

1

One should state more clearly in the docs the fact that FedDataset are regular torch dataset indexed from 0 to n-1. Each dataset's strategy's associated hyper-parameters should be directly accessible...

Dummy Dataset is nor reproducible nor flexible enough

2

DummyDataset should be seeded and the structure of the data should be customizable.