README.md

scripts_rackham

Scripts to be run on Rackham.

Workflow

0. Zip

This is local, in the nsphs_ml_qt folder:

./scripts_rackham/00_create_starter_zip.sh

Creates a file called nsphs_ml_qt_starter_zip.zip

20 start

These are run scripts that can be run with/out sbatch, as these sbatch all that needs to be sbatched

21 create

Create an experiment file and a dataset

25 run

Run an experiment file

28 analyse

Analyse the results of an experiment file

29 zip

Analyse the results of an experiment file

Old workflow

The old workflow, with numbers from 2 to and including 19 directly calls GCAE train, then project, then evaluate, without generating intermediate results.

Prefer using the new workflow.

2. Start simulated dataset 1

./nsphs_ml_qt/scripts_rackham/02_start_1.sh

10. Create simulated dataset 1

sbatch nsphs_ml_qt/scripts_rackham/10_create_dataset_1.sh

11. Train on dataset 1

sbatch nsphs_ml_qt/scripts_rackham/11_train_on_dataset_1.sh

12. Project on dataset 1

sbatch nsphs_ml_qt/scripts_rackham/12_project_on_dataset_1.sh

13. Plot on dataset 1

sbatch nsphs_ml_qt/scripts_rackham/13_plot_on_dataset_1.sh

14. Animate on dataset 1

sbatch nsphs_ml_qt/scripts_rackham/14_animate_on_dataset_1.sh

15. Evaluate on dataset 1

sbatch nsphs_ml_qt/scripts_rackham/15_evaluate_on_dataset_1.sh

98. Clean Rackham

./98_clean_rackham.sh