Create an OmicsTweezer model

Builds the encoder and predictor modules for one OmicsTweezer architecture. The encoder contains two fully connected blocks, and the predictor contains two fully connected blocks followed by a linear output layer and softmax.

Usage

omics_create_model(feature_num, celltype_num, dims, drops)

Arguments

feature_num

Integer scalar. Number of input genes.

celltype_num

Integer scalar. Number of output cell types.

dims

Integer vector of length 4 specifying hidden-layer dimensions.

drops

Numeric vector of length 4 specifying dropout probabilities.

Value

A named list with components:

encoder

Torch encoder module.

predictor

Torch predictor module returning cell-type proportions.

Details

In R torch, softmax over the feature axis of an `N x F` tensor is specified with `dim = 2`.