# Some of the updates and additions in R-4.1.0

## New Syntax

### Native Pipes

library(magrittr)
iris %>%
  filter(Species %in% c("setosa", "versicolor")) %>%
  head()
  
iris |>
  filter(Species %in% c("setosa", "versicolor")) |>
  head()

f <- function(x) x^2
g <- function(y) y*2

f(g(10))
10 %>% g() %>% f()
10 |> g() |> f()

#### Stack traces are easier to read?

## Try with something that throws an error

"test" %>% g() %>% f()
traceback()

"test" |> g() |> f()
traceback()

iris %>%
  filter(Species %in% c("setosa", "versicolor")) %>% 
  group_by(Sepal.Width) %>% 
  summarise(`Mean width` = mean(Sepal.Girth)) %>% 
  mutate(`Mean width` = round(`Mean width`, 1))
  
rlang::last_error()
  
iris |>
  filter(Species %in% c("setosa", "versicolor")) |>
  group_by(Sepal.Width) |>
  summarise(`Mean width` = mean(Sepal.Girth)) |> 
  mutate(`Mean width` = round(`Mean width`, 1))
  
rlang::last_error()

### Anonymous functions

function(x) x + 1
\(x) x + 1 

sapply(1:10, FUN = \(x) x + 1)

### Limitations of native pipes

c("vector", "matrix", "array") %>%
    grepl(pattern = "ar", x = .)

c("vector", "matrix", "array") |>
    grepl(pattern = "ar", x = .)

find_at <- function(x) grepl(pattern = "at", x = x)
c("vector", "matrix", "array") |>
  find_at()


c("vector", "matrix", "literate") |>
  {\(x) grepl(pattern = "at", x = x)}()

## Improvements & additions

### Concatenating factors works as I might expect

Try this in R-4.0 vs R-4.1. The former is not what I'd expect!

n <- 500
fkt1 <- factor(sample(letters, n, TRUE))
fkt2 <- factor(sample(LETTERS, n, TRUE))
head( 
    c(fkt1, fkt2)
)

### Improved performance for `grep()` et al on factors

nn <- 1e7
fkt <- factor(sample(letters, nn, TRUE))
system.time( 
    grepl("[k-p]", fkt) 
)

system.time( 
    gsub("[k-p]", "xx", fkt)
)

### Added `gregexec` function.

pattern <- "([[:alpha:]]+)([[:digit:]]+)"
s <- "Test: A1 BC23 DEF456"

regexec(pattern, s)
gregexec(pattern, s)

### Improved performance of duplicated() on sorted vectors

set.seed(1234)
x <- sample(1e3, 1e7, replace = TRUE)
y <- sort(x)
z <- y
## this stops z from knowing it is sorted
z[1] <- z[1]

system.time(yd <- duplicated(y))
system.time(zd <- duplicated(z))
identical(yd, zd)

### Splitting data.frames using "~"

## R-4.0.5
split(mtcars, f = mtcars$gear)

## R-4.1.0
split(mtcars, f = ~ gear)

### Locking base environment

## when would you do this?
test <- baseenv()
test$foo <- 1:10

## Extras

### The default C++ standard has been changed to C++14

### No more 32-bit Windows after R-4.1

### Translation in to Lithuanian

Tutorial at useR 2021 if this is interesting to you!

https://www.conftool.org/user2021/index.php?page=browseSessions&search=Translating+R