ADD length-of-stay stats summarization - Githubissues

tufts-ml / covid19-forecasting

Simulation software for forecasting demand at various stages of hospitalization

MIT License

3 stars 0 forks source link

ADD length-of-stay stats summarization #33

Closed michaelchughes closed 3 years ago

michaelchughes commented 3 years ago

Changes:

Adds a params.json file that matches stats in the target Table 2 from cdc docs
Updates run_forecast.py and PatientTrajectory.py so that important summary stats can be automatically produced and inspected.

2 new types of results files are produced:

1: individual simulated patient LOS results

los_results_per_patient-random_seed=1001.csv

2: aggregate summaries (means, percentiles) of LOS results across simulated patients

for all patients : los_results_summary-random_seed=1001.csv
for patients NOT on the ventilator: los_results_summary_isOnVent0-random_seed=1001.csv
for patients on the ventilator: los_results_summary_isOnVent1-random_seed=1001.csv
for patients NOT terminal: los_results_summary_isTerminal0-random_seed=1001.csv
for patients who are terminal: los_results_summary_isTerminal1-random_seed=1001.csv