Consider the following code. It seems like the errors that occur when y only 
has a key column is a bug. Yes?

DT<-data.table(a=rep(1:2,each=3),b=1:6,key="a")

y<-J(a=c(0,1),bb=c(10,11),key="a") # y, with key column and one more column
merge(y,DT) # fine
merge(y,DT,all=TRUE) # fine

y<-J(a=c(0,1),key="a") # y with only a key column
merge(y,DT) # fine
merge(y,DT,all=TRUE) # error
merge(DT,y,all=TRUE) # a different error

_______________________________________________
datatable-help mailing list
[email protected]
https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/datatable-help

Reply via email to