Don't require dplyr anywhere #1131
Conversation
Codecov Report
@@ Coverage Diff @@
## main #1131 +/- ##
==========================================
+ Coverage 92.33% 92.35% +0.01%
==========================================
Files 46 46
Lines 2650 2654 +4
==========================================
+ Hits 2447 2451 +4
Misses 203 203
|
|
This is how benchmark results would change (along with a 95% confidence interval in relative change) if 8f839e6 is merged into main:
Further explanation regarding interpretation and methodology can be found in the documentation. |
| @@ -1,16 +1,16 @@ | |||
| test_that("caching utils make right blocks with semi-colon", { | |||
| blocks_simple_uncached <- compute_parse_data_nested(c("1 + 1", "2; 1+1")) %>% | |||
| dplyr::mutate(is_cached = FALSE) %>% | |||
| mutate(is_cached = FALSE) %>% | |||
There was a problem hiding this comment.
where's %>% coming from if not dplyr?
There was a problem hiding this comment.
{magritrr}? We'll at some point transition to |> if minimal supported R version has it.
R/compat-dplyr.R
Outdated
| @@ -57,3 +57,12 @@ map_dfr <- function(.x, .f, ...) { | |||
| res <- map(.x, .f, ...) | |||
| vec_rbind(!!!res) | |||
| } | |||
|
|
|||
| mutate <- function(.data, ...) { | |||
There was a problem hiding this comment.
use base::transform() instead? use case here seems like a drop-in
…nse, !!, etc.) and it's not intended to be used in source code anyways
|
This is how benchmark results would change (along with a 95% confidence interval in relative change) if 68977f3 is merged into main:
Further explanation regarding interpretation and methodology can be found in the documentation. |
| select(token, terminal, text, newlines, spaces) | ||
| cols <- c('token', 'terminal', 'text', 'newlines', 'spaces') | ||
| pd$child[[1]][, cols] |
There was a problem hiding this comment.
maybe subset(pd$child[[1]], select = cols)?
There was a problem hiding this comment.
or since we've magrittr:
pd$child[[1]] %>%
subset(select = c(...))There was a problem hiding this comment.
I don't know... the audience of the vignette is probably R developers who know subset() but I think it's one of these functions that are not used often anymore these days... ] is not very pretty here I know but probably easier to understand for most.
Superseeds #1128.