with_seed: Random seed

In withr: Run Code 'With' Temporarily Modified Global State

Random seed

Description

with_seed() runs code with a specific random seed and resets it afterwards.

with_preserve_seed() runs code with the current random seed and resets it afterwards.

Usage

with_seed(
  seed,
  code,
  .rng_kind = NULL,
  .rng_normal_kind = NULL,
  .rng_sample_kind = NULL
)

local_seed(
  seed,
  .local_envir = parent.frame(),
  .rng_kind = NULL,
  .rng_normal_kind = NULL,
  .rng_sample_kind = NULL
)

with_preserve_seed(code)

local_preserve_seed(.local_envir = parent.frame())

Arguments

seed	⁠[integer(1)]⁠ The random seed to use to evaluate the code.
code	[any] Code to execute in the temporary environment
.rng_kind, .rng_normal_kind, .rng_sample_kind	⁠[character(1)]⁠ Kind of RNG to use. Passed as the kind, normal.kind, and sample.kind arguments of RNGkind().
.local_envir	⁠[environment]⁠ The environment to use for scoping.

Value

[any]
The results of the evaluation of the code argument.

See Also

withr for examples

Examples

# Same random values:
with_preserve_seed(runif(5))
with_preserve_seed(runif(5))

# Use a pseudorandom value as seed to advance the RNG and pick a different
# value for the next call:
with_seed(seed <- sample.int(.Machine$integer.max, 1L), runif(5))
with_seed(seed, runif(5))
with_seed(seed <- sample.int(.Machine$integer.max, 1L), runif(5))

withr documentation built on Oct. 28, 2024, 5:07 p.m.