Ivan Leung

[Ivan Leung @urganmax]

An elegant way to share variables across #rshiny modules via R6 objects, courtesy of @jiwanheo.

A practical way to learn about R6 class.

https://t.co/Ie1hu5GVz3

[Ivan Leung @urganmax]

@JNursiyono @randyboyes @mdancho84 An alternative is building a text classification model - e.g., #tidymodels - convert text descriptions to numeric features for model training.

Good intro by @juliasilge:
https://t.co/H6bnVUEZaa

[Ivan Leung @urganmax]

@JNursiyono @randyboyes @mdancho84 @JNursiyono, recommend using dplyr::case_when() - to create rules - with stringr::str_detect() - to identify keywords for categories.

The #rstudio community is a good forum to ask questions 🙂
https://t.co/GGzQ3giLxy

[Ivan Leung @urganmax]

@mdancho84 Storing list-columns in {tibbles} - e.g., mutate(data = map(col, fun)) - to keep relevant data(frame) in one place - especially useful when modeling 🙂

[Ivan Leung @urganmax]

@PhilippBayer @mdancho84 @IsabellaGhement Used to question my friend's obsession with list and lapply()...happy to say {purrr} is one of the best investment to date 🙂

Hit the gas, and we have {furrr}; thanks to @dvaughan32.

[Ivan Leung @urganmax]

@JagersbergKnut @mdancho84 {tidytable}📦combines the most useful/common verbs in {dplyr/tidyr} with {data.table} speed 🤩

Courtesy of @markfairbanks10

[Ivan Leung @urganmax]

@mdancho84 {hacksaw} 📦 by @daranzolin offers more flavours of group_split() that allows splitting by common {dplyr} verbs - e.g., filter_split(), select_split().

https://t.co/wJd6InMivo

[Ivan Leung @urganmax]

#rstats #shiny (noun) #spotfire to draw shapefile for geofencing analyses.

[Ivan Leung @urganmax]

Dev's ideal friend - persistent storage without the chains - courtesy of @CockroachDB.

[Ivan Leung @urganmax]

@IsabellaGhement #> # A tibble: 2 x 3
#> Species sepal petal
#> <chr> <list> <list>
#> 1 setosa_versicolor <tibble [100 x 2]> <tibble [100 x 2]>
#> 2 virginica <tibble [50 x 2]> <tibble [50 x 2]>

(3/3)

[Ivan Leung @urganmax]

@IsabellaGhement iris_nest |>
dplyr::filter(grepl("(?i)seto|ver", Species)) |>
dplyr::summarise(
dplyr::across(Species, paste0, collapse = "_"),
dplyr::across(tidyselect:::where(is.list),
~ list(dplyr::bind_rows(.x)))
) |>
dplyr::bind_rows(iris_virg)

(2/3)

[Ivan Leung @urganmax]

Is https://t.co/ds7S0g7qSp the right place to file a minor documentation in #RStats {base} 📦? If not, where should it be filed? 🙏🏻
@R_Contributors

[Ivan Leung @urganmax]

TIL {purrr} partial() with the `...` argument; weird syntax but it's🤩
Imagine creating many styles in {openxlsx} with a common center `halign`:

ctr <- partial(createStyle, halign = "center", ... = )
ctr(numFmt = "0.0%")
# same as
createStyle(halign = "center", numFmt = "0.0%")

[Ivan Leung @urganmax]

@tyluRp Same experience here!

Resource monitor shows it helps 🙂

[Ivan Leung @urganmax]

@MitchTanney @rstudio @_RachaelDempsey Timeless advice, courtesy of @MitchTanney 🤩

Applies to any industry, mine (energy) included. Law of nature knows no boundaries 🙂

🙏🏻 to @_RachaelDempsey @rstudio for the event.

[Ivan Leung @urganmax]

@skyetetra My go-to for binary files IO - {qs}

GH:
https://t.co/V8nVF2IZU7

[Ivan Leung @urganmax]

@joshyam_ Another option is via regular expression:

grepl("(?i)the united states", text)

The (?i) flag is for case insensitive.

[Ivan Leung @urganmax]

When {clipr} fails and you're too tired to mess with X11, {datapasta} got you covered😎

🙏 to @MilesMcBain

GH:
https://t.co/DnPBkDPTLg