Skip to contents

Creates a Graph that can be used to robustify any subsequent learner. Performs the following steps:

The graph is built conservatively, i.e. the function always tries to assure everything works. If a learner is provided, some steps can be left out, i.e. if the learner can deal with factor variables, no encoding is performed.

All input arguments are cloned and have no references in common with the returned Graph.

Usage

pipeline_robustify(
  task = NULL,
  learner = NULL,
  impute_missings = NULL,
  factors_to_numeric = NULL,
  max_cardinality = 1000,
  ordered_action = "factor",
  character_action = "factor",
  POSIXct_action = "numeric"
)

Arguments

task

Task
A Task to create a robustifying pipeline for. Optional, if omitted, the "worst possible" Task is assumed and the full pipeline is created.

learner

Learner
A learner to create a robustifying pipeline for. Optional, if omitted, the "worst possible" Learner is assumed and a more conservative pipeline is built.

impute_missings

logical(1) | NULL
Should missing values be imputed? Defaults to NULL: imputes if the task has missing values (or factors that are not encoded to numerics) and the learner can not handle them.

factors_to_numeric

logical(1) | NULL
Should (ordered and unordered) factors be encoded? Defaults to NULL: encodes if the task has factors (or character columns that get converted to factor) and the learner can not handle factors.

max_cardinality

integer(1)
Maximum number of factor levels allowed. See above. Default: 1000.

ordered_action

character(1)
How to handle ordered columns: "factor" (default) or "factor!": convert to factor columns; "numeric" or "numeric!": convert to numeric columns; "integer" or "integer!": convert to integer columns; "ignore" or "ignore!": ignore. When task is given and has no ordered columns, or when learner is given and can handle ordered, then "factor", "numeric" and "integer" are treated like "ignore". This means it is necessary to add the exclamation point to override Task or Learner properties when given. "ignore" and "ignore!" therefore behave completely identically, "ignore!" is only present for consistency.
When ordered features are converted to factor, then they are treated like factor features further down in the pipeline, and are possibly eventually converted to numerics, but in a different way: factors get one-hot encoded, ordered_action = "numeric" converts ordered using as.numeric to their integer-valued rank.

character_action

character(1)
How to handle character columns: "factor" (default) or "factor!": convert to factor columns; "matrix" or "matrix!": Use PipeOpTextVectorizer. "ignore" or "ignore!": ignore. When task is given and has no character columns, or when learner is given and can handle character, then "factor" and "matrix" are treated like "ignore". This means it is necessary to add the exclamation point to override Task or Learner properties when given. "ignore" and "ignore!" therefore behave completely identically, "ignore!" is only present for consistency.
When character columns are converted to factor, then they are treated like factor further down in the pipeline, and are possibly eventually converted to numerics, using one-hot encoding.

POSIXct_action

character(1)
How to handle POSIXct columns: "numeric" (default) or "numeric!": convert to numeric columns; "datefeatures" or "datefeatures!": Use PipeOpDateFeatures. "ignore" or "ignore!": ignore. When task is given and has no POSIXct columns, or when learner is given and can handle POSIXct, then "numeric" and "datefeatures" are treated like "ignore". This means it is necessary to add the exclamation point to override Task or Learner properties when given. "ignore" and "ignore!" therefore behave completely identically, "ignore!" is only present for consistency.

Value

Graph

Examples