azureml-automl-runtime package

Packages

column_purpose_detection

Init for column purpose detection module.

distributed

Init for distributed module.

estimation

Module for all estimators.

featurization

Init for featurization module.

transformer

Init for transformer module.

notebook_generation
onnx_convert

Init for the onnx_convert module.

sampling

Classes and methods for performing sampling.

scoring

Classes and methods for performing scoring.

shared

Init for standard AutoML modules. This package is imported directly by AutoML components in the Jasmine repo.

stats_computation

Init for Statistics computation module.

sweeping

Classes and methods for performing feature sweeping.

Modules

automl_pipeline

Wrapper class for AutoML pipeline objects.

automl_run_context

Context manager that wraps an AutoML run context.

cpu_utilities

Utility for getting CPU related information during AutoML training.

data_context

Holding the data context classes.

data_transformation

Holding the featurization functions.

dataprep_utilities

Utility methods for interacting with azureml.dataprep.

ensemble_base

Base module for ensembling previous AutoML iterations.

faults_verifier

Class for the Data faults and misconfigurations for AutoML.

feature_skus_utilities

Code to handle AutoML feature SKUs in the SDK.

featurization_info_provider

Base class for objects that provide featurization info.

fit_output

Wrapper for pipeline fit output.

fit_pipeline

Code used to fit pipeline.

frequency_fixer

The utility functions to check the

pipeline_run_helper

Helper methods to execute an AutoML pipeline fit.

preprocess

Global methods used during an AutoML fit method for pre-processing raw data into meaningful features.

constants

Metrics constants.

scoring

Computation of AutoML model evaluation metrics.

utilities

Utilities for computing model evaluation metrics.

stack_ensemble_base

Base module for ensembling previous AutoML iterations.

streaming_data_context

Holding the streaming data context classes.

streaming_pipeline_wrapper

Wrapper for the final pipeline in the streaming scenario.

timeseries

Includes classes for storing timeseries preprocessing related functions.

training_utilities

Utilities used during AutoML training.

voting_ensemble_base

Base module for ensembling previous AutoML iterations.