That's a very helpful answer. We just need to preach the gospel of stringr a little harder to all those stringi people.
1 Like
I'm of the mind that if people have already figured out the stringi API, they should just keep using it, as stringr is really just a nicer API for a subset of stringi.
Full disclosure: I mostly use base regex because it's always there, though I recognize that the stringi engine is usually faster.
But stringr works so much better with %>%!
Than base, yes; than stringi no. I mostly find myself using regex inside dplyr verbs anyway, so I need to type the column name regardless.
In some large applications I’ve found base regex to be faster than stringr, but I think it’s because of the return value. Many stringr functions return an object of length equal to (or a multiple of) the input, which can be a lot.
4 Likes
Thought this might be a good idea to try out the poll feature - TBH not sure what the min/max options do
In preview looks like a checkbox so maybe you can suggest between 1 and 5 (those are my min/max options)
Which would be preferred additions to which are loaded automatically in library(tidyverse)?
I realize there may be practical issues with regard to individual packages
- lubridate
- forcats
- stringr
- readxl
- rvest
- broom
0 voters