Awesome project btw, it was a correct decision to follow dplyr. I think you should consider overload operators for ergonomics. Most data scientists prefer using operators for doing mathematical stuff instead of calling functions named as add, divide, multiply …
Excellent language! I’m using it for scientific computing and like it a lot. There are a few but useful and good libraries. Arraymancer,ggplotnim, alea and stats. Also, great Python integration nimpy library.