Comment by political12345

Why not use R's data.table? You can achieve literally the same pipeline style flow, internally it's written in C, it's not the fastest but gets the job done.

If you want real speed but no external package, kdb/q is probably better (has q-sql).

Good work for the effort etc. but I often don't see the value added of these toy projects of people apart from personal development and a bullet point on the CV of the author.