There's something in your data that makes the model computationally
singular when you take the various subsettings... Can you provide a
small reproducible example so we can help narrow it down? It looks
like you're using different data for each mlogit though so I'm not
sure how the comparison that v2 & v5 fail while the other works is
relevant though...
Michael
On Thu, Apr 19, 2012 at 4:03 PM, geek girl <geek.girl.only at gmail.com>
wrote:> I am learning five mlogits as follows
>
> v1.model<-mlogit(v1~1|v2+v3+v4+v5, data=mlogit.v1.data,
reflevel="1")
> v2.model<-mlogit(v2~1|v1+v3+v4+v5, data=mlogit.v2.data,
reflevel="1")
> v3.model<-mlogit(v3~1|v1+v2+v4+v5, data=mlogit.v3.data,
reflevel="1")
> v4.model<-mlogit(v4~1|v1+v2+v3+v5, data=mlogit.v4.data,
reflevel="1")
> v5.model<-mlogit(v5~1|v1+v2+v3+v4, data=mlogit.v5.data,
reflevel="1")
>
> v2 and v5 give me the error below during learning, the other 3 models work
> fine
>
> "Error in solve.default(H, g[!fixed]) : system is computationally
singular:
> reciprocal condition number = 1.12239e-16"
>
> What does this error mean?
>
> Thanks
>
> ? ? ? ?[[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.