rnim VS ggplotnim

While indeed we are less people developing stuff in Nim compared to even the Julia community (which itself is of course much smaller than say Python), we do have cover a large amount of the typical needs in the scientific computing domain. And where we miss stuff it's a) easy to wrap C/C++ or b) simply call Julia, R or Python (As a personal reference I'm doing data analysis & numerical physics stuff in context of my PhD in physics and I literally do everything in Nim. The only significant C dependency {and only as a shared lib} I depend on is libhdf5 via nimhdf5).

use Rnim to access the R bindings

If you're willing to help out, you'll surely be able to do anything you need. If Nim libraries fail, you can also always call R directly from Nim via Rnim!

Maybe also of interest is a nascent package for R calling Nim (or vice versa): https://github.com/SciNim/rnim

c2nim is a tool to translate ANSI C code to Nim. The output is human-readable Nim code that is meant to be tweaked by hand after the translation process. If you are tired of wrapping C library, you can try futhark which supports "simply import C header files directly into Nim". Similar to futhark, cinterop allows one to interop with C/C++ code without having to create wrappers. nimLUA is a glue code generator to bind Nim and Lua together using Nim's powerful macro. nimpy and nimporter is a bridge between Nim and Python. rnim is a bridge between R and Nim. nimjl is a bridge between Nim and Julia! Last but not least, genny generates a shared library and bindings for many languages such as Python, Node.js, C.

I am risking to be offtopic, but there is somenthing interesting happening in Nim, where someone wrote a wrapper to R: https://github.com/SciNim/rnim

That is certainly true (and why added a general "embed plot data as bitmap into SVG/PDF" option to https://github.com/Vindaar/ggplotnim that works not only for raster heatmaps). But realistically such plots are often not ideal anyway (too many data points in a plot is often a sign that a different type of plot would be better; typically one that aggregates in some way) and it's just another argument to make the data for plots available as well.

Well, personally as I write almost all my code in Nim and am the developer of ggplotnim [0], I simply write a source code snippet with some short Nim code, generate a plot and dump the filename into the Org file.

If I had more time and wanted something more convenient and magical, I would probably write a elisp function that takes X Y (Z) columns and generates a plot from those using a simple Nim program in the back that receives the data, generates the plot and returns it somehow. Haven't given this much thought though.

[0]: https://github.com/Vindaar/ggplotnim

I have been using Nim for all of my recent data munging and analysis. There's https://github.com/Vindaar/ggplotnim for plots (among others) and everything else has just been normal code. There's also https://github.com/SciNim/Datamancer if you need something more like tidyverse.