[R-sig-ME] lmekin complains about dimnames

David Duffy David.Duffy at qimr.edu.au
Fri May 18 04:09:08 CEST 2012


On Thu, 17 May 2012, Joshua Wiley wrote:

> require(kinship2)
> gadat <- with(adat, pedigree(PersonID, FatherID, MotherID, sex,
                famid = FamilyID))
> kamat <- kinship(gadat)

> (gfit1 <- lmekin(Outcome ~ factor(sex) + (1 | FamilyID), data=adat,
> varlist=kamat))
> # Error in bdsmatrix.reconcile(varlist, bname) :
> #  No dimnames found on a variance matrix

kamat is a kinship matrix, so you need (1|id) rather than FamilyID.  But 
aside from that, there seems to be a problem in the new kinship, as it
only sets dimnames for the second dimension.  An ugly temporary solution 
is:

dimnames(kamat)[[1]] <-  dimnames(kamat)[[2]]

The second problem is that the individual IDs were non-unique. Since 
PersonIDs were non-unique, kinship generates a new ID, which is now needed 
to match up the correct kinship values

adat$gid <- paste(adat$FamilyID,adat$PersonID,sep="/")
lmekin(Outcome ~ factor(sex) + (1 | gid),
        data=adat,varlist=list(gid=kamat))


-- 
| David Duffy (MBBS PhD)                                         ,-_|\
| email: davidD at qimr.edu.au  ph: INT+61+7+3362-0217 fax: -0101  /     *
| Epidemiology Unit, Queensland Institute of Medical Research   \_,-._/
| 300 Herston Rd, Brisbane, Queensland 4029, Australia  GPG 4D0B994A v



More information about the R-sig-mixed-models mailing list