Ok, phew. Then it's the same issue I asked on stack overflow, and Simon Ubanek helped me out today. Will fix. It's the new code for fast character sorting.
> It is the source straight from SVN. I've also found it randomly > segfaulting on me today, and have not been able to reproduce how. But > an hour back on the CRAN version and I haven't hit a segfault. > > On 15 November 2011 15:23, Matthew Dowle <[email protected]> wrote: >> Is that with 1.7.3 on R-Forge? If so I think I know. >> >>> I have a data.frame DF with a factor column "foo", it has X unique >>> levels. When I do >>> DT=data.table(DF) >>> key(DT) = "foo" >>> >>> DT then has X-1 unique levels, and one repeated level. I then >>> continually get messages saying that repeated levels are no longer >>> allowed. >>> >>> The level that goes missing looks like: >>> >>> "blahblah�0123456789" >>> >>> >>> There is a special character that I don't know what it is in there I >>> hope it shows through in the email. >>> After setting key(DT) , there is a completely different level that is >>> duplicated, and the level above goes missing. >>> >>> >>> I'm not quite sure how to replicate this with data I can show. It >>> seems a weird bug.  Any thoughts? >>> _______________________________________________ >>> datatable-help mailing list >>> [email protected] >>> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/datatable-help >> >> >> > _______________________________________________ datatable-help mailing list [email protected] https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/datatable-help
