composer.datasets.brats#

composer.datasets.brats

Functions

coin_flip

composer.datasets.brats.coin_flip

dataclass

Returns the same class as was passed in, with dunder methods added based on the fields defined in the class.

get_data_split

composer.datasets.brats.get_data_split

get_split

composer.datasets.brats.get_split

load_data

composer.datasets.brats.load_data

random_augmentation

composer.datasets.brats.random_augmentation

Classes

Blur

composer.datasets.brats.Blur

Brightness

composer.datasets.brats.Brightness

Contrast

composer.datasets.brats.Contrast

Crop

composer.datasets.brats.Crop

DataLoader

Protocol for custom DataLoaders compatible with torch.utils.data.DataLoader.

Flips

composer.datasets.brats.Flips

Noise

composer.datasets.brats.Noise

PytTrain

composer.datasets.brats.PytTrain

PytVal

composer.datasets.brats.PytVal

Transpose

composer.datasets.brats.Transpose

Hparams

These classes are used with yahp for YAML-based configuration.

BratsDatasetHparams

Defines an instance of the BraTS dataset for image segmentation.

DataloaderHparams

Hyperparameters to initialize a Dataloader.

DatasetHparams

Abstract base class for hyperparameters to initialize a dataset.

Attributes

  • Dataset

  • PATCH_SIZE

class composer.datasets.brats.BratsDatasetHparams(is_train=True, drop_last=True, shuffle=True, datadir=None, oversampling=0.33)[source]#

Bases: composer.datasets.hparams.DatasetHparams

Defines an instance of the BraTS dataset for image segmentation.

Parameters

  • datadir (str) – The path to the data directory.

  • is_train (bool) – Whether to load the training data (the default) or validation data.

  • drop_last (bool) – If the number of samples is not divisible by the batch size, whether to drop the last batch (the default) or pad the last batch with zeros.

  • shuffle (bool) – Whether to shuffle the dataset. Defaults to True.

  • oversampling (float) – The oversampling ratio to use.

initialize_object(batch_size, dataloader_hparams)[source]#

Creates a DataLoader or DataloaderSpec for this dataset.

Parameters

  • batch_size (int) – The size of the batch the dataloader should yield. This batch size is device-specific and already incorporates the world size.

  • dataloader_hparams (DataloaderHparams) – The dataset-independent hparams for the dataloader

Returns

  • Dataloader or DataSpec – The dataloader, or if the dataloader yields batches of custom types,

  • a :class:`DataSpec`.