pylastmlextension package¶

Subpackages¶

Write basic model class for training interface and method to write a json configuration file for the model.

class pylastmlextension.base.BaseModel(name, feature_schema)[source]¶

Bases: ABC

Base class for all models based on lightgbm.

save(output_dir)[source]¶

Save the model and config json file to output_dir.

save_model(output_dir)[source]¶

Save the model to output_dir.

Convert the model to a dictionary.

train(x, y, weights=None)[source]¶

Train the model.

Pre-defined schema for pylast datalevels

This model contains two quantile regression models to estimate the sigma - quantile_model_16: 16th percentile - quantile_model_84: 84th percentile

class pylastmlextension.sigma_estimator.SigmaEstimator(name, feature_schema, n_trials, cv=3)[source]¶