Efficient Value-based Subset Methods on (Long) Vectors

The slicev_ - methods are similar to the sb_ - methods, except they don't require an indexing vector, and are designed for memory efficiency.

counv(y, v, from, to) counts how often a value, or range of values, v, occurs in a vector subset y[from:to].

Usage

slicev_x(x, ...)

# Default S3 method
slicev_x(
  x,
  ...,
  y = x,
  v = NULL,
  na = FALSE,
  r = TRUE,
  from = NULL,
  to = NULL,
  use.names = TRUE,
  sticky = getOption("squarebrackets.sticky", FALSE)
)

slicev_set(x, ...)

# Default S3 method
slicev_set(
  x,
  ...,
  y = x,
  v = NULL,
  na = FALSE,
  r = TRUE,
  from = NULL,
  to = NULL,
  rp,
  tf
)

countv(y, ..., v = NULL, na = FALSE, r = TRUE, from = NULL, to = NULL)

Arguments

x: an atomic vector.
For slicev_set() it must be a mutable_atomic variable.
...: See squarebrackets_slicev.
y, v, na, r: See squarebrackets_slicev.
from, to: see cp_seq.
use.names: Boolean, indicating if flat names should be preserved.
Note that, since the slicev_ methods operates on flat indices only, dimensions and dimnames are always dropped.
sticky: see squarebrackets_options.
rp, tf: see squarebrackets_modify.

Value

Similar to the sb_ methods.

For countv(): A single number, giving the number of elements matching the specified condition.

Examples


# basic idea ====
nms <- c(letters, LETTERS, month.abb, month.name) |> rep_len(1e6)
x <- mutable_atomic(1:1e6, names = nms)
head(x)
#> a b c d e f 
#> 1 2 3 4 5 6 
#> mutable_atomic 
#> typeof:  integer 

# memory efficient form of sum(x <= 10):
countv(x, v = c(-Inf, 10))
#> [1] 10

# extract all elements of x with the name "a":
slicev_x(x, y = names(x), v = "a") |> head()
#>   a   a   a   a   a   a 
#>   1  77 153 229 305 381 
#> mutable_atomic 
#> typeof:  integer 

# find all x smaller than or equal to 5, and replace with `-1000`:
slicev_set(x, y = x, v = c(-Inf, 5), rp = -1000L)
head(x, n = 10)
#>     a     b     c     d     e     f     g     h     i     j 
#> -1000 -1000 -1000 -1000 -1000     6     7     8     9    10 
#> mutable_atomic 
#> typeof:  integer 


################################################################################
# Numeric range ====
#
x <- mutable_atomic(1:1e6)
head(x)
#> [1] 1 2 3 4 5 6
#> mutable_atomic 
#> typeof:  integer 
slicev_x(x, v= c(-Inf, 5)) # x[x <= 5]
#> [1] 1 2 3 4 5
#> mutable_atomic 
#> typeof:  integer 


################################################################################
# Character ====
#
x <- stringi::stri_rand_shuffle(rep("hello", 1e5))
head(x)
#> [1] "loelh" "oellh" "oelhl" "lehlo" "leolh" "eohll"
slicev_x(x, v = "hello") |> head() # find "hello"
#> [1] "hello" "hello" "hello" "hello" "hello" "hello"

# find 2 possible misspellings of "hello":
slicev_x(x, v = c("holle", "helol")) |> head()
#> [1] "holle" "holle" "helol" "helol" "holle" "helol"