immuneML.workflows.steps package

Subpackages

• immuneML.workflows.steps.data_splitter package
  • Submodules
  • immuneML.workflows.steps.data_splitter.DataSplitter module
    • DataSplitter
      • DataSplitter.k_fold_split()
      • DataSplitter.leave_one_out_stratification_split()
      • DataSplitter.loocv_split()
      • DataSplitter.manual_split()
      • DataSplitter.random_split()
      • DataSplitter.run()
      • DataSplitter.stratified_k_fold_split()
  • immuneML.workflows.steps.data_splitter.DataSplitterParams module
    • DataSplitterParams
      • DataSplitterParams.dataset
      • DataSplitterParams.label_config
      • DataSplitterParams.paths
      • DataSplitterParams.split_config
      • DataSplitterParams.split_count
      • DataSplitterParams.split_strategy
      • DataSplitterParams.training_percentage
  • immuneML.workflows.steps.data_splitter.LeaveOneOutSplitter module
    • LeaveOneOutSplitter
      • LeaveOneOutSplitter.split_dataset()
  • immuneML.workflows.steps.data_splitter.ManualSplitter module
    • ManualSplitter
      • ManualSplitter.split_dataset()
  • immuneML.workflows.steps.data_splitter.Util module
    • Util
      • Util.make_dataset()
      • Util.prepare_path()
  • Module contents

Submodules

immuneML.workflows.steps.DataEncoder module

class immuneML.workflows.steps.DataEncoder.DataEncoder

Bases: Step

static run(input_params: StepParams = None)

immuneML.workflows.steps.DataEncoderParams module

class immuneML.workflows.steps.DataEncoderParams.DataEncoderParams(dataset: immuneML.data_model.datasets.Dataset.Dataset, encoder: immuneML.encodings.DatasetEncoder.DatasetEncoder, encoder_params: immuneML.encodings.EncoderParams.EncoderParams)

Bases: StepParams

dataset: Dataset

encoder: DatasetEncoder

encoder_params: EncoderParams

immuneML.workflows.steps.MLMethodAssessment module

class immuneML.workflows.steps.MLMethodAssessment.MLMethodAssessment

Bases: Step

fieldnames = ['run', 'optimal_method_params', 'method', 'encoding_params', 'encoding', 'evaluated_on']

static run(input_params: MLMethodAssessmentParams = None)

immuneML.workflows.steps.MLMethodAssessmentParams module

class immuneML.workflows.steps.MLMethodAssessmentParams.MLMethodAssessmentParams(method: MLMethod, dataset: Dataset, metrics: set, optimization_metric: ClassificationMetric, label: Label, path: Path, split_index: int, predictions_path: Path, ml_score_path: Path)

Bases: StepParams

immuneML.workflows.steps.MLMethodTrainer module

class immuneML.workflows.steps.MLMethodTrainer.MLMethodTrainer

Bases: Step

static run(input_params: MLMethodTrainerParams = None)

static store(method: MLMethod, input_params: MLMethodTrainerParams)

immuneML.workflows.steps.MLMethodTrainerParams module

class immuneML.workflows.steps.MLMethodTrainerParams.MLMethodTrainerParams(method: ~immuneML.ml_methods.classifiers.MLMethod.MLMethod, dataset: ~immuneML.data_model.datasets.Dataset.Dataset, result_path: ~pathlib.Path, label: <module 'immuneML.environment.Label' from '/Users/milenpa/PycharmProjects/BMIimmuneML/immuneML/environment/Label.py'>, model_selection_cv: bool, model_selection_n_folds: int, cores_for_training: int, train_predictions_path: ~pathlib.Path, optimization_metric: str)

Bases: StepParams

immuneML.workflows.steps.SignalImplanter module

immuneML.workflows.steps.Step module

class immuneML.workflows.steps.Step.Step

Bases: object

This class encapsulates steps in the analysis which will likely be often used, such as:

• dataset encoding

• training of machine learning models

• signal implanting in repertoires without any signals etc.

For a custom analysis which is not likely to be repeated for different settings (e.g. such as with a different encoding), create a custom class inheriting AbstractProcess from workflows.processes package.

abstract static run(input_params: StepParams = None)

immuneML.workflows.steps.StepParams module

class immuneML.workflows.steps.StepParams.StepParams

Bases: object

Module contents