Closed santiagcarcal closed 2 years ago
Unlike the GLM method, the method anova.ergm()
does not have the test
argument. It's primary job is model comparison. What exactly do you want to achieve?
Edit: I think I now see the discrepancies in the documentation. Thank you for reporting!
THanks for your answer. I want to know how much variation in link formation is explained by each attribute.
I am fitting a valued ERGM model of a bipartite network. I am using
sum
as the valued term and the weights of the bipartite matrix as the response variable. I'm using packagesergm
andergm.count
to fit the model andnetwork
to construct the objects. I've fitted the model and obtained the estimates, but I want to call an ANOVA table for the model usinganova(ergm.1, test="Chi")
. But I get this error:Error: $ operator is invalid for atomic vectors
. Alternatively when I useanova(ergm.1)
I only get an 'empty' anova table, as if I had not specified the variables.