biolord.BiolordClassifyModule

class biolord.BiolordClassifyModule(categorical_attributes_missing=None, classify_all=False, logits=False, bias=True, classification_penalty=0.1, classifier_penalty=0.0001, classifier_nn_width=128, classifier_nn_depth=2, classifier_dropout_rate=0.1, loss_regression='normal', **kwargs)

The biolord-classify module.

A BiolordModule accompanied by regressors for ordered classes and classifiers for categorical classes.

Parameters:

• categorical_attributes_missing (Optional[dict[str, str]]) – Categorical categories representing un-labeled cells.

• classify_all (bool) – Whether to classify all classes or only semi-supervised classes.

• logits (bool) – Classifier output type.

• bias (bool) – Whether to add bias to the regressor.

• classification_penalty (float) – Classification penalty strength.

• classifier_nn_width (int) – Classifier’s layer width.

• classifier_nn_depth (int) – Classifier’s number of layers.

• classifier_dropout_rate (float) – Classifier’s dropout rate.

• loss_regression (Literal['normal', 'mse']) – Loss function for regressors

• kwargs (Any) – Keyword arguments for BiolordModule.

Attributes table

Methods table

classify(genes)	Run classification.
loss(tensors, inference_outputs, ...)	Compute the loss.

Attributes

Methods

classify

BiolordClassifyModule.classify(genes)

Run classification.

Parameters:

genes (Tensor) – Gene expression used for classification.

Return type:

dict[str, Tensor]

Returns:

Classification output, probability for each ordered attribute and the regression value for ordered attributes

loss

BiolordClassifyModule.loss(tensors, inference_outputs, generative_outputs)

Compute the loss.

Parameters:

• tensors (dict[str, Tensor]) – Considered model inputs.

• inference_outputs (dict[Literal['latent_unknown_attributes'], Tensor]) – Inference step outputs.

• generative_outputs (dict[Literal['distribution', 'means', 'variances'], Tensor]) – Generative step outputs.

Return type:

dict[str, float]

Returns:

The loss elements.