epimodel / epimodelcovid

Network-Based Epidemic Modeling of Transmission Dynamics for SARS-CoV-2

EpiModelCOVID

Modules for simulating SARS-CoV-2 transmission dynamics in different epidemiological settings, developed as an extension to our general network-based epidemic modeling platform, EpiModel.

EpiModel and EpiModelCOVID use the statistical framework of temporal exponential-family random graph models to fit and simulate models of dynamic contact networks. These statistical methods have been developed and implemented as open-source software, building on the extensive efforts of the Statnet research group to build software tools for the representation, analysis, and visualization of complex network data.

These packages combine these Statnet methods with an individual-based epidemic modeling engine to simulate SARS-Cov-2 transmission over networks, allowing for complex dependencies between the network, epidemiological, and demographic changes in the simulated populations. Readers new to these methods are recommended to consult our EpiModel resources, including our main methods paper Vignette describing the theory and implementation.

Installation

You can install EpiModelCOVID in R using remotes:

install.packages("EpiModel", dependencies = TRUE)
remotes::install_github("EpiModel/EpiModelCOVID")

Documentation on using this software package is forthcoming, although limited function documentation is provided within the package and available with the help(package = "EpiModelCOVID") command.

epimodelcovid's Issues

EpiModelCOVID is not currently compatible with EpiModel@multinets changes

For example, it calls init_tergmLite, which has been removed in the multinets branch.

I will spend some time reviewing this package and updating initialization/simulation as appropriate for pending core EpiModel changes.

del$dx referenced but not set

In mod_infection.R, infect_covid_corporate() and infect_covid_contacttrace() reference del$dx but that variable is never set (it does not exist).

# Case isolation with diagnosed or symptomatic infection
        if (at >= act.rate.dx.inter.time) {
          del$actRate[del$dx == 2] <- del$actRate[del$dx == 2] *
                                      act.rate.dx.inter.rr
        }

Bias in ergm network resimulation using TERGM API

related to EpiModel/EpiModelHIV-p#191

Error with nwstats saving

Parallel issue to https://github.com/EpiModel/EpiModelHIV-p/issues/141 for EpiModelCOVID

Error in update_cumlative_edgelist with ContactTrace branch

Running the test script for Kristin's EpiModelCOVID, I am encountering this error:

	A ERROR occured in module 'initialize.FUN' at step `Initialization Step` (1)
Error: Can't recycle input of size 4 to size 5.
Run `rlang::last_error()` to see where the error occurred.

within update_cumlative_edgelist specifically at the line:

  terminated_edges <- is.na(el_cuml[["current"]]) & is.na(el_cuml[["stop"]])
  if (any(terminated_edges)) {
    el_cuml[terminated_edges, ][["stop"]] <- at - 1
  }

Relevant files for reproducing the error:

1. Script: https://github.com/EpiModel/COVID-ContactTrace/blob/updatePackages/R/03-epi-model-calib.R (use this updatePackages branch for now)
2. Package: https://github.com/EpiModel/EpiModelCOVID/tree/ContactTrace

Fix nwstats calculation to use modern approach

See implementation in corporate moduleset.

Updates needed for init and simnet modules for ContactTrace

Remove bespoke calc_nwstats functions in package.

Looking for clarification on "discord_edgelist_covid_ship" function

Hello Sam,
This is Nisarg, part of the group adapting your cruise code for the hospital context. I was looking for your help to provide an insight into why "nw = 1" in line 221 of "mod-infection.R" , within the "discord_edgelist_covid_ship" function?

I noticed it did not seem to affect the outcome for the patient/ healthcare worker parameters (different parameter values for # of patients, health care workers & rooms compared to the cruise ship code) if I change it to "nw" or "nw=2" since the infection function is dependent on the network selected on the discordant edge.*

Thanks a lot for your help,
Nisarg

*edit: changing it to nw or nw=2 did have an effect on the cruise ship code. limiting the simulation output to both crew and passengers & only passengers respectively.

Mismatch in positional and unique IDs using get_partners() function

When filtering for eligible IDs for case investigation (e.g., status in c(a,ic,ip)), noticing that these are not always translating into the IDs being pulled by the get_partners() for the discordant edgelist.

For example: my idsEligCI are: 15, 657, 759, and 835. The index IDs in the discordant edgelist are 15, 659, 761, and 837. Seems like this is an issue with the positional versus unique IDs.

Relevant code for this issue: [https://github.com/EpiModel/EpiModelCOVID/blob/ContactTrace/R/mod-intervention.R]

Lines 27-28
idsEligCI <- which(active == 1 & status %in% c("a", "ic", "ip") & dxStatus == 2 & is.na(eligible.case))
Line 45
del_ct <- get_partners(dat, idsEligCI)

Global function definitions

In running R CMD check on ContactTrace, we are getting this as a NOTE:

   resim_nets_covid_contacttrace: no visible global function definition
     for ‘%n%<-’
   resim_nets_covid_contacttrace: no visible global function definition
     for ‘%n%’
   resim_nets_covid_contacttrace: no visible global function definition
     for ‘as.edgelist’
   resim_nets_covid_corporate: no visible global function definition for
     ‘%n%<-’
   resim_nets_covid_corporate: no visible global function definition for
     ‘%n%’
   resim_nets_covid_corporate: no visible global function definition for
     ‘as.edgelist’
   resim_nets_covid_ship: no visible global function definition for
     ‘%n%<-’
   resim_nets_covid_ship: no visible global function definition for ‘%n%’
   resim_nets_covid_ship: no visible global function definition for
     ‘as.edgelist’
   Undefined global functions or variables:
     %n% %n%<- as.edgelist layer

Is this something we get also with EpiModelHIV? Is it an issue with imports?