[statnet_help] Error of nodesqrtcovar terms in valued ERGM

Pavel N. Krivitsky p.krivitsky at unsw.edu.au
Wed Dec 23 15:19:00 PST 2020


Dear Yu-Lun,

Indeed, it looks like a bug in the term definition. Thanks for
reporting it! You can get a fix by installing from the GitHub
repository, though you may be able to get it to work before the next
release by evaluating the following code:

InitWtErgmTerm.nodeisqrtcovar<-function (nw, arglist, response, ...) {
### Check the network and arguments to make sure they are appropriate.
a <- check.ErgmTerm(nw, arglist, directed = TRUE, nonnegative=TRUE, response=response,
varnames = c(),
vartypes = c(),
defaultvalues = list(),
required = c())
### Process the arguments

list(name="nodeisqrtcovar",
coef.names = "nodeisqrtcovar",
dependence = TRUE,
# arithmetic mean >= geometric mean
minval = 0
)
}


I hope this helps,
Pavel

On Wed, 2020-12-23 at 15:42 -0500, Yu-Lun Chen wrote:

> Dear all,

>

> Good day! I am fitting directed, valued ERGMs for relatively small

> social networks (12 nodes), where I experience issues

> with nodeisqrtcovar and nodeosqrtcovar terms. I intended to model

> Individual heterogeneity for both out-and in-degrees, yet I wasn't

> able to include both terms in the model. The error reads: 

>

> Error in check.ErgmTerm(nw, arglist, directed = TRUE, nonnegative =

> TRUE,  :

>   all_identical(c(length(varnames), length(vartypes),

> length(defaultvalues),  .... is not TRUE

>

>  Instead, when I tried either term or the undirected term

> (nodesqrtcovar), the model worked out fine, only that

> nodesqrtcovar tended to show convergence issues. Is it possible to

> fit nodeisqrtcovar and nodeosqrtcovar terms simultaneously? Thank you

> so much!

>

> Happy holidays,

> Yu-Lun

>

>

> _______________________________________________

> statnet_help mailing list

> statnet_help at u.washington.edu

> http://mailman13.u.washington.edu/mailman/listinfo/statnet_help


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman13.u.washington.edu/pipermail/statnet_help/attachments/20201224/867d821b/attachment.html>


More information about the statnet_help mailing list