Possible regression in stan 1.1.*

[Guy Freeman]

Dear all,

I just thought I should let you know that the attached model works without a problem on rstan 1.0.3 installed through R 2.15.3 on Windows 7 (64-bit version), but not with rstan 1.1.1. By "not working" I mean the model does not compile when using set_cppo("debug"), and does compile with set_cppo("fast") but when run makes Windows freeze up by using up all available RAM on 64-bit R (except with an extremely small dataset) or crashing under 32-bit R. If there are any more details I can or should provide (e.g. messages when crashing) please let me know.

Guy

PS. A "#340" error forced me to change the extension of the stan model file.

[Bob Carpenter]

Thanks. There was a memory leak in our
speedup for multivariate normal that we've
tracked down and fixed.

We hope to release 1.2.0 tomorrow or Wednesday and
it will contain the leak-free version.

Sorry about that.

- Bob

> --
> You received this message because you are subscribed to the Google Groups "stan users mailing list" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to stan-users+...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>

[Guy Freeman]

Wonderful, thank you Bob for your superlative work, as ever. Is there a public list of current issues/bugs?

Guy

[Daniel Lee]

Yes. Our current TO-DO list can be found here:
https://github.com/stan-dev/stan/blob/master/TO-DO.txt

We also use github's issue tracker for bugs in Stan:
https://github.com/organizations/stan-dev/dashboard/issues

If there are things on the list you think you could help with, feel free to contribute.

Daniel

[Bob Carpenter]

Guy implicitly brings up the excellent point -- we
should have a list of known bugs in the current version.

As is, we've been removing them from the issue tracker
as closed when we fix them on the master branch.

If we closed them only when we release a version of Stan
with a patch, it'd be easier for users to see what the known
issues are.

- Bob

> stan-users+...@__googlegroups.com.
> > For more options, visit https://groups.google.com/__groups/opt_out <https://groups.google.com/groups/opt_out>.

> >
> >
>
> --
> You received this message because you are subscribed to the Google Groups "stan users mailing list" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to

> stan-users+...@googlegroups.com <mailto:stan-users%2Bunsu...@googlegroups.com>.