Hacker News Reader

I avoid joins in R like a plague. Been burned. If I am working with a finite number of colummns, I just use base R match and add column by column. yes it breaks DRY but I don't get weird surprises. and I can re-run any section of the code at any time from any point without breaking things or joining the same and creating unintended duplicate columns

View on HN

What kind of weird surprises have you run into? Are they specific to R?

Take a look at the powerjoin package. It let's you specify strict rules for your joins, for example requiring unique keys in the right-side table when doing a left join. It's got some other neat features as well.

Reply on news.ycombinator.com