In UchidaMizuki/partialised: Enhanced Adverbial Functions

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)

adverbial (former partialised)

adverbial provides new_partialised() and new_composed(), which extend partial() and compose() functions of purrr to make it easier to extract and replace arguments and functions, and has additional adverbial functions such as as_step() for step-by-step data processing.

Installation

You can install the development version of adverbial from GitHub with:

# the released version from CRAN:
install.packages("adverbial")

# the development version from GitHub:
# install.packages("devtools")
devtools::install_github("UchidaMizuki/adverbial")

Examples

library(adverbial)

Enhanced partialised functions

new_partialised() is an enhanced version of partial() from purrr. It allows you to extract and replace arguments of the function.

dist <- function(x, y) {
  sqrt(x ^ 2 + y ^ 2)
}

pdist <- new_partialised(
  dist,
  list(x = 3)
)
pdist
pdist(y = 4)

# Get partialised arguments
pdist[]
pdist$x
pdist$y

pdist$x <- 6
pdist(y = 8)

pdist$y <- 8
pdist()

Enhanced composed functions

new_composed() is an enhanced version of compose() from purrr. It allows you to extract and replace functions in the composition.

square <- function(x) x^2
cdist <- new_composed(
  list(
    square = square,
    sum = sum,
    sqrt = sqrt
  ),
  dir = "forward"
)
cdist
cdist(1:10)

# Get composed functions
cdist[]
cdist$sum <- new_partialised(sum, list(na.rm = TRUE))

cdist(c(1:10, NA))

Step-by-step data processing

step_by_step() defines a step-by-step data processing pipeline by passing a character vector with step names and descriptions.

as_step() converts an existing function into a step function that can be used in a pipeline. Generated functions check if a step is correct for objects created with step-by-step() and act as a normal function for other objects. With as_step(f) (without passing a second argument) you can add another function to step-by-step data processing.

end_step() is a function that can be used to end the step-by-step data processing pipeline and return the result.

# Define a step-by-step data processing pipeline
dist_calculator <- step_by_step(c(
  square_step = "Square the input",
  sum_step = "Sum the squares",
  sqrt_step = "Take the square root"
))

# Define the steps
square_step <- as_step(function(x) x^2, "square_step")
sum_step <- as_step(sum, "sum_step")
sqrt_step <- as_step(sqrt, "sqrt_step")

square_step
sum_step
sqrt_step

dist <- dist_calculator(c(1:10, NA))
dist

dist <- dist |> 
  square_step() |> 
  sum_step(na.rm = TRUE) |>
  sqrt_step()
dist
end_step(dist)