PyProphet

class pyprophet.scoring.pyprophet.PyProphet(config: RunnerIOConfig)

Bases: object

Orchestrates the semi-supervised learning and scoring workflow.

config

The configuration object for the workflow.

semi_supervised_learner

The semi-supervised learner instance.

__init__(config: RunnerIOConfig)

__weakref__

list of weak references to the object (if defined)

_apply_weights_on_exp(experiment, weights)

Applies weights to the experiment and updates the learner.

Parameters:

• experiment – The experiment data.

• weights – The pre-trained weights.

Returns:

The updated learner instance.

Return type:

object

_build_result(table, final_classifier, score_columns, experiment)

Builds the final result object after scoring and error estimation.

Parameters:

• table – The input data table.

• final_classifier – The trained classifier.

• score_columns – List of score column names.

• experiment – The experiment data.

Returns:

Result object, scorer instance, and classifier table.

Return type:

tuple

_learn(experiment, score_columns)

Trains the semi-supervised learner using randomized folds.

Parameters:

• experiment – The experiment data.

• score_columns – List of score column names.

Returns:

The trained learner instance.

Return type:

object

_setup_experiment(table)

Prepares the experiment data by scaling features and logging a summary.

Parameters:

table – The input data table.

Returns:

Prepared experiment and list of score columns.

Return type:

tuple

apply_weights(table, loaded_weights)

Applies pre-trained weights to the input data.

Parameters:

• table – The input data table.

• loaded_weights – The pre-trained weights.

Returns:

Result object, scorer instance, and classifier table.

Return type:

tuple

learn_and_apply(table)

Performs learning and scoring on the input data.

Parameters:

table – The input data table.

Returns:

Result object, scorer instance, and classifier table.

Return type:

tuple