torch_ecg.components

This module contains the components for training and evaluating models.

Input classes

InputConfig(*args, input_type, n_channels[, ...])	A Class to store the configuration of the input.
WaveformInput(config)	Waveform input.
FFTInput(config)	Inputs from the FFT, via concatenating the amplitudes and the phases.
SpectrogramInput(config)	Inputs from the spectro-temporal domain.

Output classes

ClassificationOutput(*args, **kwargs)	Class that maintains the output of a (typically single-label) classification task.
MultiLabelClassificationOutput(*args, **kwargs)	Class that maintains the output of a multi-label classification task.
SequenceTaggingOutput	alias of SequenceTaggingOutput
SequenceLabellingOutput(*args, **kwargs)	Class that maintains the output of a sequence tagging task.
WaveDelineationOutput(*args, **kwargs)	Class that maintains the output of a wave delineation task.
RPeaksDetectionOutput(*args, **kwargs)	Class that maintains the output of an R peaks detection task.

Loggers

LoggerManager([log_dir, log_suffix])

Manager of loggers.

Metrics

ClassificationMetrics([multi_label, macro, ...])	Metrics for the task of classification.
RPeaksDetectionMetrics([thr, extra_metrics])	Metrics for the task of R peaks detection, as proposed in CPSC2019.
WaveDelineationMetrics([macro, tol, ...])	Metrics for the task of ECG wave delineation.

Trainer

BaseTrainer(model, dataset_cls, ...[, ...])

Abstract base class for trainers.