see the thread on stackoverflow: [optimizing a C# function from R](http://stackoverflow.com/questions/17214741/is-it-possible-to-call-a-r-statistics-function-to-optimize-c-sharp-function).
I am interested in supporting this. First: is this actually already doable. If not, likely that I can implement by similarity with rClr.
Comments: Hi, The context in the post is C# calling R calling back a C# function. Indeed one thing that comes to mind is to use rClr in the R statements to set this up, but I suspect this would end up badly. As I recall there can be only one CLR per process, same as only one R engine per process, so I doubt this is technically trivial. It may be possible for rClr to hook-up to an existing CLR, but right now it is not the behavior I think. Even if it ran OK, things could go quickly wrong with "reentrant calls" in case where error handling comes into play. Still worth exploring as a learning experience, but I reckon a solution with R.NET itself is preferable in the long term. Cheers
I am interested in supporting this. First: is this actually already doable. If not, likely that I can implement by similarity with rClr.
Comments: Hi, The context in the post is C# calling R calling back a C# function. Indeed one thing that comes to mind is to use rClr in the R statements to set this up, but I suspect this would end up badly. As I recall there can be only one CLR per process, same as only one R engine per process, so I doubt this is technically trivial. It may be possible for rClr to hook-up to an existing CLR, but right now it is not the behavior I think. Even if it ran OK, things could go quickly wrong with "reentrant calls" in case where error handling comes into play. Still worth exploring as a learning experience, but I reckon a solution with R.NET itself is preferable in the long term. Cheers