Assesser: Assesser Object

In signaux-faibles/MLsegmentr: Model evaluation on data segments

Description

This is an object used to assess one or several models, on the whole data or specific segments, for one or several target variables.

Usage

Assesser

Format

[R6::R6Class] object.

Fields

* 'test_data' :: 'data.frame()'
Data used for defining targets, segments, optionnaly predictions, features and others.

* 'id_names' :: 'character()'
Names of one or several columns of the test data that uniquely define each row

* 'evaluation_funs' :: '[eval_function]'
EvaluationFunction that will be used to evaluated the models.

S3 Methods

* For all three next methods, several inputs are given as lists, and recycled with a warning if lenghts are not equal.

* 'set_predictions()', 'set_segments()', 'set_targets()' :: 'character() => invisible(self)'
defines the columns storing respectively predictions, segments and targets. 'target_type' and 'prediction_type' are derived from the column names.

* 'assess_model'
VOID -> any, ...
Assess the models thanks to the eval_function object. Additional ... arguments are transmitted to the Evaluation function.

signaux-faibles/MLsegmentr documentation built on Aug. 29, 2019, 2:22 p.m.