What is your favorite R function you've learned this 2024? Mine: I've loved understanding !! and enquo from rlang. #rstats - ThreadSky

leonardohansa.com • 73 days ago

What is your favorite R function you've learned this 2024?

Mine: I've loved understanding !! and enquo from rlang.
#rstats

Comments

jeremy-data.bsky.social•71 days ago

for me, it was card() from bslib https://rstudio.github.io/bslib/articles/cards/index.html#hello-card #Shiny

leonardohansa.com•71 days ago

Great tip! I use bslib whenever I create a Shiny app and didn't know about cards.

Thanks a lot for sharing! Now I have new ingredients to try 💙

tjmahr.com•73 days ago

posterior::rvar()

mackaszechno.bsky.social•73 days ago

Interesting...
https://mc-stan.org/posterior/articles/rvar.html

leonardohansa.com•71 days ago

Didn't know about this one but I'm already loving it!

mackaszechno.bsky.social•73 days ago

...the rest aside, my favourite was learning ain and amatch from the stringdist library that lets you fuzzy match strings.

leonardohansa.com•71 days ago

Wow, great tip! I can really use this when getting data from some clients.

I've used other packages in the past but I like how straightforward these functions are.

mackaszechno.bsky.social•73 days ago

{{ var }} for me, which, correct me if I'm wrong, is kinda the same thing.

domleoni.bsky.social•73 days ago

I am (also?) confused by {{ vs the above.

tjmahr.com•73 days ago

f({{ x }}) is a shortcut for a common pattern of bottling up a name with
xq <- enquo(x)

and injecting it with later

f(!!xq)

mackaszechno.bsky.social•73 days ago

See link. For me, I mostly use it in functions to pass column names to tidyverse verbs inside the function as arguments of the function.
https://rlang.r-lib.org/reference/embrace-operator.html

apsteinmetz.bsky.social•73 days ago

That is a dream of mine, because I still don't.

mackaszechno.bsky.social•73 days ago

*Refers to own notes*
# The injection operators {{, !!, and !!! are an extension of the
# R syntax developed for tidyverse packages.
# !! plus enquo() OR the embrace operator: {{ }}
# !!! is used to splice a list/vector of some sort into arguments of some quoting expression.

elfrank7.bsky.social•73 days ago

There are sooo many, but `grepl()` for pattern matching has helped me a lot!!

mackaszechno.bsky.social•73 days ago

See Danielle Navarro's excellent blog on regex. Time well spent.
https://bsky.app/profile/djnavarro.net/post/3ldkxevtlys2o

elfrank7.bsky.social•73 days ago

Cooool thanks for the share!

kurtstat.bsky.social•72 days ago

I learned a bit more about cut() to help me create frequency distributions. Can't help but think there must be a better way, though...

leonardohansa.com•71 days ago

I used cut() quite a lot until I found out about findInterval(). A bit less intuitive but much faster -at least for what I needed

matanhakim.bsky.social•72 days ago

The `.by` argument in {dplyr} functions, makes much more sense when summarising ang mutating.

leonardohansa.com•71 days ago

Don't you like how explicit group_by() is, compared with the too-compact data.table style?

matanhakim.bsky.social•70 days ago

I don't like the implicit "peeling" in `summarise`, sometimes when you forget the grouping weird things happen.

Also, I feel that the grouping is mostly an auxiliary to the other verb, so it makes more sense for me as an argument.

Finally, I don't find it compact or implicit, so it works for me 💯

leonardohansa.com•69 days ago

I agree on the auxiliary nature of group_by(). In fact, this makes it the most difficult dplyr verb to introduce to my students, because its behavior is unique.

Reading you, I might stop using group_by() -even though I find it intuitive, it might simplify my classes. I hadn't thought about it 😅

Comments

Posting Rules

Reply