loss_function_graph: Loss function regularised by the graph information

In ff98li/scRGNet: Perform scRNA-seq Analysis using Single-cell Graphical Neural Network (scGNN) Framework

Description

Reconstruction + KL divergence losses summed over all elements and batch

Usage

loss_function_graph(
  recon_x,
  x,
  mu = NULL,
  logvar = NULL,
  graph_regu = NULL,
  gamma_param = 1,
  regu_mat = NULL,
  regu_type = NULL,
  regu_param = 0.001,
  model = "AE",
  reduction = "sum"
)

Arguments

recon_x	Reconstructed scRNA-seq matrix
x	Original input scRNA-seq matrix
mu	mu
logvar	logvar
graph_regu	graph_regu
gamma_param	gamma_param
regu_mat	reg_mat
regu_type	Type of regulariser
regu_param	reg_param
model	AE
reduction	sum #' @references \insertRefscGNNscRGNet

ff98li/scRGNet documentation built on Jan. 14, 2022, 4:58 a.m.