4 ms·
Strange to see R as dreaded... It's not a tool for everything, but for the right niche (data analysis), it seems like a very sensible choice. Could it be that
by tompazourek 5y ago
Strange to see R as dreaded... It's not a tool for everything, but for the right niche (data analysis), it seems like a very sensible choice.
Could it be that it's something that non-programmers are forced to use if they want to do some advanced stuff? And that these non-programmers dread programming in R not because of the language, but because it's programming.
If not, then what is so dreadful about R?
- fwip 5y agoR is mostly just a different paradigm than most programmers are used to. A surprising number of them are scared off by R's 1-based indexing by default, or the slightly-unfamiliar syntax. It's designed for REPL usage foremost, and with vectors / dataframes as the normal unit of operation. There's also a whole bunch of built-in functions with unfamiliar names and options, and R libraries tend to be similarly opaque to the user with a programming (non-stats) background. For an R-like written for programmers, see Julia.
- disgruntledphd2 5y ago> For an R-like written for programmers, see Julia. Yeah, if you look at the people behind dataframes.jl, it's a lot of people who were involved in R in the earlier days. And there is no better environment for solving a data analysis problem than R. Python can work better for some DS use-cases (sklearn is ace), and it's definitely better for glue but if you need to understand what the hell is going on for a product/service/whatever, R is definitely the right choice.
- dunefox 5y ago"Sensible choice due to the ecosystem and tools" and "loving the language for its features" are two very different things.