FixedGroupsPipeline: Chain multiple transformations with different argument values...

In rearrr: Rearranging Data

Chain multiple transformations with different argument values per group

Description

\Sexpr[results=rd, stage=render]{lifecycle::badge("experimental")}

Build a pipeline of transformations to be applied sequentially.

Specify different argument values for each group in a fixed set of groups. E.g. if your data.frame contains 5 groups, you provide 5 argument values for each of the non-constant arguments (see `var_args`).

The number of expected groups is specified during initialization and the input `data` must be grouped such that it contains that exact number of groups.

Transformations are applied to groups separately, why the given transformation function only receives the subset of `data` belonging to the current group.

Standard workflow: Instantiate pipeline -> Add transformations -> Apply to data

To apply the same arguments to all groups, see Pipeline.

To apply generated argument values to an arbitrary number of groups, see GeneratedPipeline.

Super class

rearrr::Pipeline -> FixedGroupsPipeline

Public fields

transformations

list of transformations to apply.

names

Names of the transformations.

num_groups

Number of groups the pipeline will be applied to.

Methods

Public methods

• FixedGroupsPipeline$new()

• FixedGroupsPipeline$add_transformation()

• FixedGroupsPipeline$apply()

• FixedGroupsPipeline$print()

• FixedGroupsPipeline$clone()

---

Method new()

Initialize the pipeline with the number of groups the pipeline will be applied to.

Usage

FixedGroupsPipeline$new(num_groups)

Arguments

num_groups

Number of groups the pipeline will be applied to.

---

Method add_transformation()

Add a transformation to the pipeline.

Usage

FixedGroupsPipeline$add_transformation(fn, args, var_args, name)

Arguments

fn

Function that performs the transformation.

args

Named list with arguments for the `fn` function.

var_args

Named list of arguments with list of differing values for each group.

E.g. list("a" = list(1, 2, 3), "b" = list("a", "b", "c")) given 3 groups.

By adding ".apply" with a list of TRUE/FALSE flags, the transformation can be disabled for a specific group.

E.g. list(".apply" = list(TRUE, FALSE, TRUE), ....

name

Name of the transformation step. Must be unique.

Returns

The pipeline. To allow chaining of methods.

---

Method apply()

Apply the pipeline to a data.frame.

Usage

FixedGroupsPipeline$apply(data, verbose = FALSE)

Arguments

data

data.frame with the same number of groups as pre-registered in the pipeline.

You can find the number of groups in `data` with `dplyr::n_groups(data)`. The number of groups expected by the pipeline can be accessed with `pipe$num_groups`.

verbose

Whether to print the progress.

Returns

Transformed version of `data`.

---

Method print()

Print an overview of the pipeline.

Usage

FixedGroupsPipeline$print(...)

Arguments

...

further arguments passed to or from other methods.

Returns

The pipeline. To allow chaining of methods.

---

Method clone()

The objects of this class are cloneable with this method.

Usage

FixedGroupsPipeline$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.

Author(s)

Ludvig Renbo Olsen, [e-mail]

See Also

Other pipelines: GeneratedPipeline, Pipeline

Examples

# Attach package
library(rearrr)
library(dplyr)

# Create a data frame
# We group it by G so we have 3 groups
df <- data.frame(
  "Index" = 1:12,
  "A" = c(1:4, 9:12, 15:18),
  "G" = rep(1:3, each = 4)
) %>%
  dplyr::group_by(G)

# Create new pipeline
pipe <- FixedGroupsPipeline$new(num_groups = 3)

# Add 2D rotation transformation
pipe$add_transformation(
  fn = rotate_2d,
  args = list(
    x_col = "Index",
    y_col = "A",
    suffix = "",
    overwrite = TRUE
  ),
  var_args = list(
    degrees = list(45, 90, 180),
    origin = list(c(0, 0), c(1, 2), c(-1, 0))
  ),
  name = "Rotate"
)

# Add the `cluster_group` transformation
# As the function is fed an ungrouped subset of `data`,
# i.e. the rows of that group, we need to specify `group_cols` in `args`
# That is specific to `cluster_groups()` though
# Also note `.apply` in `var_args` which tells the pipeline *not*
# to apply this transformation to the second group
pipe$add_transformation(
  fn = cluster_groups,
  args = list(
    cols = c("Index", "A"),
    suffix = "",
    overwrite = TRUE,
    group_cols = "G"
  ),
  var_args = list(
    multiplier = list(0.5, 1, 5),
    .apply = list(TRUE, FALSE, TRUE)
  ),
  name = "Cluster"
)

# Check pipeline object
pipe

# Apply pipeline to already grouped data.frame
# Enable `verbose` to print progress
pipe$apply(df, verbose = TRUE)

rearrr documentation built on April 4, 2025, 1:07 a.m.