catboost.get_object_importance {catboost}R Documentation

Calculate the object importances

Description

Calculate the object importances (see https://catboost.ai/docs/concepts/ostr.html). This is the implementation of the LeafInfluence algorithm from the following paper: https://arxiv.org/pdf/1802.06640.pdf

Usage

catboost.get_object_importance(model, pool, train_pool, top_size = -1,
  type = "Average", update_method = "SinglePoint", thread_count = -1,
  ostr_type = NULL)

Arguments

model

The model obtained as the result of training.

Default value: Required argument

pool

The pool for which you want to evaluate the object importances.

Default value: Required argument

train_pool

The pool on which the model has been trained.

Default value: Required argument

top_size

Method returns the result of the top_size most important train objects. If -1, then the top size is not limited.

Default value: -1

update_method

Description of the update set methods are given in section 3.1.3 of the paper.

Possible values:

  • 'SinglePoint'

  • 'TopKLeaves' It is posible to set top size : TopKLeaves:top=2.

  • 'AllPoints'

Default value: 'SinglePoint'

thread_count

The number of threads to use when applying the model. If -1, then the number of threads is set to the number of CPU cores.

Allows you to optimize the speed of execution. This parameter doesn't affect results.

Default value: -1

ostr_type

Deprecated parameter, use 'type' instead.

type.

Possible values:

  • 'Average'

    Method returns the mean train objects scores for all input objects.

  • 'PerObject'

    Method returns the train objects scores for every input object.

Default value: 'Average'

See Also

https://catboost.ai/docs/concepts/r-reference_catboost-get_object_importance.html


[Package catboost version 0.23 Index]