This package leverages the heavy lifting done by dcmdata to then fit various CDA models to the simulated data. The intention is to create wrappers around each model type so that the final script which executes the model can be much cleaner/simpler.
Currently there is support for the following models (in various states of effectiveness!):
- RUM model using explicit priors in JAGS
- Reduced-RUM model in JAGS
- Reparamaterized-DINA model in JAGS
- Reparamaterized-DINA model in STAN
sim <- rDINAJagsSim(data, q, jagsModel="RDINA.jags", adaptSteps = 1000, burnInSteps = 1000, numSavedSteps = 10000, thinSteps = 1)
Simulation work is hard enough without having to worry about keeping track of which version I simulated etc.
Use devtools to install from the github repo.
Test run as part of a CI environment using testthat.
Dave Rackham [email protected]
MIT License (see License.txt)