Sensitivity Dispatcher Examples

The sensitivity dispatcher is a tool for running one or more predefined sensitivity experiments using the dataset notebooks/double-entry-data/double_entry_final.csv. It is intended to use when running experiments on a server. This page gives some example commands to run with it.

See also

The Sensitivity Dispatcher Usage documentation.

Example 1:

python scripts/sensitivity_dispatcher.py --max_processes 4 --categories region_holdout

Masks data for each of the 41 regions, conditions the default model on the masked data and saves predicted infection course of held out region to sensitivity_default/region_holdout/{held out region}.pkl (note that with 41 runs, this takes a very long time).

Example 2:

python scripts/sensitivity_dispatcher.py --max_processes 4 --dry_run --categories npi_leaveout cases_threshold

Prints the commands that would be run for the npi_leaveout and cases_threshold sensitivity experiments.

Output:

Running Univariate Sensitivity Analysis
---------------------------------------

Categories: ['npi_leaveout', 'cases_threshold']
You have requested 14 runs
Performing Dry Run
python scripts/sensitivity_analysis/npi_leaveout.py --model_type default --n_samples 2000 --n_chains 4 --exp_tag npi_leaveout --npis 0
python scripts/sensitivity_analysis/npi_leaveout.py --model_type default --n_samples 2000 --n_chains 4 --exp_tag npi_leaveout --npis 1
python scripts/sensitivity_analysis/npi_leaveout.py --model_type default --n_samples 2000 --n_chains 4 --exp_tag npi_leaveout --npis 2
python scripts/sensitivity_analysis/npi_leaveout.py --model_type default --n_samples 2000 --n_chains 4 --exp_tag npi_leaveout --npis 3
python scripts/sensitivity_analysis/npi_leaveout.py --model_type default --n_samples 2000 --n_chains 4 --exp_tag npi_leaveout --npis 4
python scripts/sensitivity_analysis/npi_leaveout.py --model_type default --n_samples 2000 --n_chains 4 --exp_tag npi_leaveout --npis 5
python scripts/sensitivity_analysis/npi_leaveout.py --model_type default --n_samples 2000 --n_chains 4 --exp_tag npi_leaveout --npis 6
python scripts/sensitivity_analysis/npi_leaveout.py --model_type default --n_samples 2000 --n_chains 4 --exp_tag npi_leaveout --npis 7
python scripts/sensitivity_analysis/npi_leaveout.py --model_type default --n_samples 2000 --n_chains 4 --exp_tag npi_leaveout --npis 8
python scripts/sensitivity_analysis/npi_leaveout.py --model_type default --n_samples 2000 --n_chains 4 --exp_tag npi_leaveout --npis 6 7
python scripts/sensitivity_analysis/preprocessing_tests.py --model_type default --n_samples 2000 --n_chains 4 --exp_tag cases_threshold --smoothing 1 --deaths_threshold 10 --cases_threshold 10
python scripts/sensitivity_analysis/preprocessing_tests.py --model_type default --n_samples 2000 --n_chains 4 --exp_tag cases_threshold --smoothing 1 --deaths_threshold 10 --cases_threshold 50
python scripts/sensitivity_analysis/preprocessing_tests.py --model_type default --n_samples 2000 --n_chains 4 --exp_tag cases_threshold --smoothing 1 --deaths_threshold 10 --cases_threshold 150
python scripts/sensitivity_analysis/preprocessing_tests.py --model_type default --n_samples 2000 --n_chains 4 --exp_tag cases_threshold --smoothing 1 --deaths_threshold 10 --cases_threshold 200