Matthew Fidler

Ah; OK; ``` >Matrix(LETTERS[1:4], 2, byrow = TRUE) > x DenseMatrix of dim 2 x 2 [A, B] [C, D] ``` So the operations for `symengine` would be the `R`...

The additional `openMp` parallelization would be minimal, but possibly worthwhile? Anyhow, thanks for the information. I will let you decide.

What about `lambertW`?

And `erf`, though it is a transformation of `pnorm`

Congrats @Marlin-Na ; Indeed that is a honor. Good luck with the CRAN release; My offer still stands (if you need any help).

In RxODE, we did factoring by symengine -> expand -> dsl parsing. It is possible, but it would be done outside of symengine.

It is multivariate polynomial factoring. However, in principle it is the same as univariate polynomial factoring applied to each term in a polynomial or non-polynomial expression. I used it for...

Not for my purposes. You are right, of course.

Hm. It is buried in the package (and it is off-topic for this repository) First after using symengine's `expand`, I split it with `rxSplitPlusQ` for the `+` and `-` operators:...

``` r library(RxODE); lapply(rxSplitPlusQ("a*exp(b+c)+d*log(e-f)-g*f"),function(x){eval(parse(text=sprintf("rxSplitPlusQ(quote(%s),mult=TRUE)",x)))}) #> [[1]] #> [1] "a" "exp(b + c)" #> #> [[2]] #> [1] "d" "log(e - f)" #> #> [[3]] #> [1] "-g" "f" ``` Created...