leifeld / btergm

Temporal Exponential Random Graph Models by Bootstrapped Pseudolikelihood
16 stars 10 forks source link

Package ‘btergm’ was removed from the CRAN repository #34

Closed Whissi closed 1 year ago

Whissi commented 1 year ago

From https://cran.r-project.org/web/packages/btergm/index.html:

Package ‘btergm’ was removed from the CRAN repository.

Formerly available versions can be obtained from the [archive](https://cran.r-project.org/src/contrib/Archive/btergm).

Archived on 2023-03-31 as requires archived package 'speedglm'.

A summary of the most recent check results can be obtained from the [check results archive](https://cran-archive.r-project.org/web/checks/2023/2023-03-31_check_results_btergm.html).

Please use the canonical form https://cran.r-project.org/package=btergm to link to this page.
Whissi commented 1 year ago

It was removed because it depends on speedglm package which was removed on 2023-04-01 for policy violation.

According to https://github.com/cole-trapnell-lab/monocle3/issues/652#issuecomment-1494760286, maintainer of speedglm package is aware of the problem and package should be back soon.

leifeld commented 1 year ago

Thanks. Yes, I was notified by CRAN and given two weeks to remove the speedglm dependency. But that was one day after I went on vacation, and when I came back, btergm was no longer on CRAN. I'm still working through my email backlog after coming back, but will hopefully get to this problem soon.

leifeld commented 1 year ago

I have made speedglm optional and resubmitted to CRAN. It should be there in a few days, but in the meantime, check the new instructions on the landing page for installation details from GitHub. speedglm is now used by default if available. If not, the btergm function uses glm instead, which is slower and takes more memory. To get a faster version, you can also try the argument usefastglm = TRUE to use the fastglm package for estimation. That's probably the fastest one, though speedglm may be more memory-efficient because it can use sparse matrices from the Matrix package.