I was able to estimate the following temporal, bipartite ERGM using the xergm
package in R:
require(xergm)
model1 <- btergm(observed_network ~ edges + b2star(2:3), R = 1000)
summary(model1)
When I went to estimate the goodness-of-fit (GOF) for the model, the following error appeared:
gof(model1)
Error in MHproposal.formula(constraints, arguments =
control$MCMC.prop.args, :
The combination of class (c), model constraints (), reference measure
(Bernoulli), and proposal weighting (default) is not implemented. Check
your arguments for typos.
My hunch is that the previous error relates to a problem with the way I specified the model for the goodness-of-fit. I used the data.table
package to construct the adjacency matrix, so it may be that xergm
and data.table
are not working well together.
After reading the XERGM and ERGM manuals, and some significant googling, I was unable to get much of a lead on the substantive meaning of the error.
Any help would be appreciated.
Aucun commentaire:
Enregistrer un commentaire