Pavel N. Krivitsky

Results 123 issues of Pavel N. Krivitsky

fixes statnet/network#86

… and eval.parent() rather than passing arguments through by name, thus preserving missing() information. @CarterButts , is this OK? A simpler alternative is to remove all arguments to `network()` except...

We seem to have missed it during the last round.

At the moment, methods for `set.vertex.attribute()` for subclasses of `network` such as those used in `networkLite` and `ergm.multi` cannot use the in-place assignment semantics of the `network` itself because the...

Using `ergm.multi`'s `N()` operator as backend for TERGM CMLE automagically enabled specifying time-varying models. These are now documented, but tests and perhaps examples need to be added.

enhancement
help wanted

Since I am primarily using `egor` as a data storage and manipulation tool, having it also depend on `igraph`, `tidygraph`, as well as `network` means that a lot of unnecessarily...

Currently, subsetting `egor` also subsets the design, which is good, but it also needs to adjust the design when, for example, a clustering variable is modified in-place (such as to...

bug

**Describe the bug** As of this writing, `kable %>% row_spec(row = 0, ...)` modifies the header of the table, but if the header has multiple rows (via `add_header_above()`), the changes...

This is to parallel the eponymous argument to `row_spec()`. I have a preliminary implementation and would be happy to submit a PR. **To Reproduce** ```` --- output: pdf_document --- ```{r}...

I've added [compatibility code](https://github.com/statnet/ergm/blob/ad708276a802ae441e6432f2205648f3139813c9/R/ergm.pl.R#L53-L57) for the upcoming release, but please switch to passing `ergm_state` objects.