Information on Model Run/Code Layout

[kzscisoft]

I am part of a consortium which is developing other COVID models, we are looking to use this particular model as a way of cross-validating these.

I noticed some of the scripts are not actually used when running the main command and there are a few references to Dropbox files which have not been included. As such I was wondering if anyone working on this project could help me gain an understanding of the workflow/structure of the code. This information would prove very valuable when it comes to applying the model to our own data/parameters.

Thank you for your time,

Kristian

[coursera-ricardor]

Some references are available in a closed issue.

https://github.com/cmmid/covid-uk/issues/1#issuecomment-607580383

I'm facing a problem with:

# Build C++ code
packageStartupMessage("Attaching C++ code...")
sourceCpp(paste0(cm_path_, "/model_v1/corona.cpp"), 
          rebuild  = cm_force_rebuild_,
          cacheDir = cm_build_dir_,
          verbose  = cm_build_verbose_)
sourceCpp(paste0(cm_path_, "/fit_v1/fit.cpp"),
          rebuild  = cm_force_rebuild_, 
          cacheDir = cm_build_dir_,
          verbose  = cm_build_verbose_)

In my windows with rtools when trying to compile covidm/model_v1/corona.cpp

In my case,I received a report for DR Congo, the geography it is quite peculiar and the infrastructure different than UK, Europe or America. Reviewing the code there is a section at the run_scenarios.R

#' set up paramaters
--
  | if(hirisk_prop_isolated > 0){
  | params <- params_set[[2]]
  | #assign population actually   isolated in high-risk
  | params$pop[[1]][["size"]] <- params$pop[[1]][["size"]] + (1-hirisk_prop_isolated)*params$pop[[2]][["size"]]
  | params$pop[[2]][["size"]] <- params$pop[[2]][["size"]]*(hirisk_prop_isolated)
  | } else {
  | params <- params_set[[1]]
  | }

So maybe it is possible to setup to consider:

Some of the cities require "River Crossing" no bridges and it is possible to view different exposure risks between countries, like:

or maybe run the model selecting some cities and not all the population in one big bucket, not possible to know in that pdf report.

name	census_198	estimate20	calculatio	annual_gro
Kinshasa	2 664 309	7273947	9463749	3.34
Lubumbashi	564 830	1283380	1786397	4.22
Mbuji-Mayi	486 235	1213726	1680991	4.16
Kananga	298 693	720362	1061181	4.96
Kisangani	317 581	682599	935977	4.02
Bukavu	167 950	471789	806940	6.94
Tshikapa	116 016	366503	587548	6.08
Kolwezi	416 122	456446	453147	-0.09
Likasi	213 862	367219	447449	2.5
Goma	77 908	249862	431342	7.06
Kikwit	149 296	294210	397737	3.84
Uvira	74 432	235136	378736	6.14
Bunia	59 598	230625	366126	5.95
Mbandaka	137 291	262814	345663	3.48
Matadi	138 798	245862	306053	2.78
Kabinda	24 789	126723	219154	7.09
Butembo	73 312	165333	217625	3.49
Mwene-Ditu	94 560	170786	195622	1.71
Isiro	78 268	147524	182900	2.72
Kindu	66 812	135534	172321	3.05

In some countries, the infrastructure could work like a shield for some communities, but in some cases the patients simply won't be able to reach a medical facility.