[Rsiena-help] Error in x$randomSeed : $ operator is invalid for atomic vectors

[Guido Conaldi]

Dear Ruth,

I am trying to fit a Siena model on three bipartite (134 x 719) networks
using RSIena rev.84 . I use the same commands I used already for several
similar models with bipartite networks, however this time all I get is
an error.

This is the stripped down version of the model, only the networks and
the defaults effects. The matrices seem fine to me. I can send you the
objects and the files in case you need them.

Thank you in advance.

Regards,
Guido

epinet511 <-
as.matrix(read.table("project_periodTest_sociomatrix_siena_1.txt",header=T))

epinet512 <-
as.matrix(read.table("project_periodTest_sociomatrix_siena_2.txt",header=T))

epinet513 <-
as.matrix(read.table("project_periodTest_sociomatrix_siena_3.txt",header=T))


dim(epinet511)
[1] 134 719

dim(epinet512)
[1] 134 719

dim(epinet513)
[1] 134 719

table(epinet511)
epinet511
    0     1 
96035   311 

table(epinet512)
epinet512
    0     1 
95972   374 

table(epinet513)
epinet513
    0     1 
95489   857 


act <- sienaNodeSet(134, nodeSetName="act")

ev <- sienaNodeSet(719, nodeSetName="ev")


foo.model <- sienaModelCreate(useStdInits = TRUE, projname =
'foo_model')

foo.coll <- sienaNet(array(c(epinet511,epinet512,epinet513),dim=c(134,
719, 3)), nodeSet=c("act", "ev"))
 
foo.data <- sienaDataCreate(foo.coll,nodeSets=list(act,ev))

foo.effects <- getEffects(foo.data)

print01Report(foo.data, foo.effects, modelname = 'foo_model')

foo.ans <- siena07(foo.coll, data=foo.data, effects=foo.effects,
batch=TRUE, verbose=TRUE)
Error in x$randomSeed : $ operator is invalid for atomic vectors
Error in exitfn() : could not find function "tkdestroy"