vim exiter

@rabaath@fosstodon.org Candidates could code in R or python. Most chose python. Yet I never saw two python solutions done in the same way. There was always a mix of weird kludges. In R, tidyverse or data.table, (or SQL, say) there's only one way to code the solution. So I'm curious what's the best "steel man" pandas solution to multiple aggregates in a group by, with arbitrary functions, and you control the names of the output columns, etc. 2/2

@rabaath@fosstodon.org this is great, thank you. I used to run a tech screen interview at my old company, the first part of which required candidates to basically do a group-by and then provide two or three different summaries. (In your example, consider computing the mean and standard deviation of amount for each country, say.) 1/

@Barredo golf clap.

@JasonAnthonyGuy I'm bummed because my very academic book was published after the download cutoff, and I don't qualify, even though it's on libgen. 3K would have doubled my royalties.

@Posit@fosstodon.org respectfully, no.

@forrestbrazeal@hachyderm.io the puffy vest is a nice touch.

@tvaziri@mastodon.social you'll make a fine little helper.