covidsims - Simulating COVID-19 dynamics in Brazil

To run the preprocessing pipeline:

python preprocess.py ../configs/T01/SP/preprocess_T01_SP.cfg

Config file must encode:

Parameter	Description
`PARAM_SOURCEPATH`	directory where the raw datafile is stored, as a list
`PARAM_TARGETPATH`	directory where results will be stored, as a list
`PARAM_DATAFILE`	the name of the raw datafile
`PARAM_DATAFIELDS`	list with the fields that correspond to region, state, city, date, #newCases, and #newDeaths, using the strings in the header of `PARAM_DATAFILE`
`PARAM_TERRITORY`	list of areas that compose the territory under analysis (i.e., territorial units)
`PARAM_POPSIZES`	dictionary of population sizes per territorial unit
`PARAM_OUTCOMES`	list of (`p`, `mu`, `rsd`) tuples, with `p` as the share of cases with some outcome category (e.g., cases with mild to moderate progression), `mu` is the mean recovery time of cases (in days), and `rsd` is the relative standard deviation of recovery time
`PARAM_MA_WINDOW`	the number of days to consider in the moving average of new cases and deaths
`PARAM_CORE_MODEL`	`Peddireddy` of `IB-forward`. See comments in ./code/sharedDefs.py/createBoL(...)
`PARAM_MASK_ERRORS`	If True, adjustments in the R(t) estimator to avoid negative values will be performed

Several UTF-8, flat files with tab-separated fields are created:

File	Description
`daily_changes.csv`	File with daily changes of the variables of interest, namely ∆S(t), ∆C(t), ∆I(t), ∆R(t), and ∆D(t)
`surveillance.csv`	File with daily status of the variables of interest, namely S(t), C(t), I(t), R(t), and D(t)
`violations.csv`	File listing the violations of quality criteria for the generated data. See criteria in ./code/sharedDefs.py/playBoL(...)
`roulette.csv`	File describing the roulette employed to assign the recovery time to an arbitrary case

The resulting data can be visualised with Excel; a template is available at ./assets/Helper_SP.xlsx

andreplima / covidsims Goto Github PK