iterate: Iterate Function Over All Combinations of User-Specified...

In dvmisc: Convenience Functions, Moving Window Statistics, and Graphics

Description

Same idea as purrr::pmap, but with some different functionality. It can runs all combinations of vector-valued arguments in ... or the 1st set, 2nd set, and so forth, and multiple trials can be run for each scenario, which can be useful for simulations.

Usage

iterate(f, ..., all_combinations = TRUE, fix = NULL, trials = 1,
  varnames = NULL)

Arguments

f	A function.
...	Arguments to f, any of which can be vector-valued.
all_combinations	Logical value for whether to iterate over all combinations of arguments in ..., or just use the first set of elements, then the second, and so on.
fix	List of arguments to f to hold fixed rather than loop over.
trials	Numeric value.
varnames	Character vector of names for values that f returns, to avoid generic labels (V1, V2, ...).

Value

Data frame.

Examples

# Define function to generate data from N(mu, sigsq) and perform t-test.
f <- function(n = 100, mu = 0, sigsq = 1, alpha = 0.05) {
  x <- rnorm(n = n, mean = mu, sd = sqrt(sigsq))
  fit <- t.test(x = x, alpha = alpha)
  return(list(t = fit$statistic, p = fit$p.value))
}

# Call f once for various sample sizes and means
f %>% iterate(n = c(100, 500), mu = c(0.1, 0.25))

# Run 100 trials for each scenario and calculate empirical power
f %>% iterate(n = c(100, 500), mu = c(0.1, 0.25), trials = 100) %>%
  group_by(n, mu) %>%
  summarise(mean(p < 0.05))

dvmisc documentation built on Dec. 18, 2019, 1:35 a.m.