splitted_mutate: splitted_mutate

In MarcellGranat/granatlib: Marcell's Personal R Package

splitted_mutate

Description

Splits the data.frame into separeta parts and evaluates it one-by-one. Useful for saving memmory during large computations.

Usage

splitted_mutate(.data, ..., split_number = 100, .keep = NULL, .progress = TRUE)

Arguments

.data	Data frame.
...	arguments to pass mutate(...).
split_number	Number of parts to split the data.frame before evaluation.
.keep	Variables to keep after eval the expression. Original columns are keept automaticly.
cores	Number of cores to use. If cores = 1 (default) then evaluation is made by simple for loop.

Value

.data overview_print

Examples

iris_samples <- tibble(data = replicate(100, sample_n(iris, 50), simplify = FALSE))

iris_samples

iris_samples %>%
 splitted_mutate(
   fit = map(data, lm, formula = Sepal.Length ~ Sepal.Width),
   coefs = map(fit, "coefficients"),
   coef1 = map_dbl(coefs, 1),
   coef2 = map_dbl(coefs, 2),
   .keep = c("coef1", "coef2") # single core
 )

iris_samples %>%
 splitted_mutate(
   fit = map(data, lm, formula = Sepal.Length ~ Sepal.Width),
   coefs = map(fit, "coefficients"),
   coef1 = map_dbl(coefs, 1),
   coef2 = map_dbl(coefs, 2),
   .keep = c("coef1", "coef2"),
   cores = 6 # multicore evaluation
 )

MarcellGranat/granatlib documentation built on July 9, 2023, 6:08 a.m.