evalml.pipelines.MulticlassClassificationPipeline.new

MulticlassClassificationPipeline.new(parameters, random_seed=0)
Constructs a new instance of the pipeline with the same component graph but with a different set of parameters.

Not to be confused with python’s __new__ method.

Parameters
  • parameters (dict) – Dictionary with component names as keys and dictionary of that component’s parameters as values. An empty dictionary or None implies using all default values for component parameters. Defaults to None.

  • random_seed (int) – Seed for the random number generator. Defaults to 0.

Returns

A new instance of this pipeline with identical components.