archetypes.torch.BiAA

class archetypes.torch.BiAA(k, m, n, device='cpu')

Biarchetype analysis implemented in PyTorch.

Parameters:

k: tuple

The number of archetypes to use for each dimension.

m: int

The number of observations in the first dimension.

n: int

The number of observations in the second dimension.

device: str

The device to use for training the model. Defaults to “cpu”.

Attributes:

A

A coefficients matrix.

B

B coefficients matrix.

C

C coefficients matrix.

D

D coefficients matrix.

Z

The archetypes matrix.

Methods

train(data, n_epochs[, learning_rate])

Train the model.

property A

A coefficients matrix.

Returns:

torch.Tensor

property B

B coefficients matrix.

Returns:

torch.Tensor

property C

C coefficients matrix.

Returns:

torch.Tensor

property D

D coefficients matrix.

Returns:

torch.Tensor

property Z

The archetypes matrix.

Returns:

torch.Tensor

train(data, n_epochs, learning_rate=0.01)

Train the model.

Parameters:

data: torch.Tensor

The data to be used for training.

n_epochs: int

The number of epochs to train the model for.

learning_rate: float

The learning rate to use for training.