Davis Vaughan

^ I seem to be hardwired to use `@return` https://cs.github.com/r-lib/clock?q=%40return, I only somewhat recently heard from Hadley that he prefers `@returns`

I made some repos for you to reproduce it again. https://github.com/DavisVaughan/testpakremote1 https://github.com/DavisVaughan/testpakremote2 - testpakremote2 has a Remote for `DavisVaughan/testpakremote1@bar` - You can do `pak::pak("DavisVaughan/testpakremote2")` to install the dev version of...

Yea, somewhat minimal reprex: ``` r library(gh) library(vctrs) x List of 4 #> $ method : chr "GET" #> $ response :List of 27 #> ..$ server : chr "GitHub.com"...

That would look something like: ``` r library(gh) library(vctrs) library(tidyr) x

Trying to convert it directly into a `SEXP` and return it is unstable. It returns sometimes, but occasionally locks up the R session or crashes it. So calling something like...

Going through the stack of: `SEXP -> rarray -> xarray -> argsort -> rarray -> SEXP` works but obviously it would be better if this was fixed. i.e. all of...

I know that - `as` is for `SEXP -> C++ type` - `wrap()` is for `C++ type -> SEXP` as summarized nicely in the first code block here http://dirk.eddelbuettel.com/code/rcpp/Rcpp-extending.pdf But...

Yea, both of these work: ```cpp // [[Rcpp::depends(xtensor)]] // [[Rcpp::plugins(cpp14)]] #include // [[Rcpp::export(rng = false)]] Rcpp::RObject rray__exp_explicit(const xt::rarray& x) { xt::rarray res = xt::exp(x); return Rcpp::RObject((SEXP) res); } // [[Rcpp::export(rng...

Small bump to have this fixed if you get the time. I was planning on doing an example that required `stddev()` to work.

Ah this looks heavily related to the prepend / append of new dimension conversation in #57. Where you have `a[1,2,2]=1` (Python and xtensor convention), in the R world we would...