That may be true for elaborate analyses, but doesn’t really address exploratory analysis that changes dramatically from command to command. The use case for REP

That may be true for elaborate analyses, but doesn’t really address exploratory analysis that changes dramatically from command to command. The use case for REPL-driven data science is experimentation, not performance