pipe_replace_step: Replace pipeline step

In pipeflow: Lightweight, General-Purpose Data Analysis Pipelines

Replace pipeline step

Description

Replaces an existing pipeline step.

Usage

pipe_replace_step(
  pip,
  step,
  fun,
  params = list(),
  description = "",
  group = step,
  keepOut = FALSE
)

Arguments

pip	Pipeline object
step	string the name of the step. Each step name must be unique.
fun	function or name of the function to be applied at the step. Both existing and anonymous/lambda functions can be used. All function parameters must have default values. If a parameter is missing a default value in the function signature, alternatively, it can be set via the params argument (see Examples section with mean() function).
params	list list of parameters to set or overwrite parameters of the passed function.
description	string optional description of the step
group	string output collected after pipeline execution (see function pipe_collect_out()) is grouped by the defined group names. By default, this is the name of the step, which comes in handy when the pipeline is copy-appended multiple times to keep the results of the same function/step grouped at one place.
keepOut	logical if FALSE (default) the output of the step is not collected when calling pipe_collect_out() after the pipeline run. This option is used to only keep the results that matter and skip intermediate results that are not needed. See also function pipe_collect_out() for more details.

Value

returns the Pipeline object invisibly

See Also

pipe_add()

Examples

p <- pipe_new("pipe", data = 1)
pipe_add(p, "add1", \(x = ~data, y = 1) x + y)
pipe_add(p, "add2", \(x = ~data, y = 2) x + y)
pipe_add(p, "mult", \(x = 1, y = 2) x * y, keepOut = TRUE)
pipe_run(p) |> pipe_collect_out()
pipe_replace_step(p, "mult", \(x = ~add1, y = ~add2) x * y, keepOut = TRUE)
pipe_run(p) |> pipe_collect_out()
try(pipe_replace_step(p, "foo", \(x = 1) x))   # step 'foo' does not exist

pipeflow documentation built on April 3, 2025, 10:50 p.m.