initialize_params: Initialize parameters for data integration or transfer

In pengminshi/cFIT: cFIT (common Factor Space Integration & Transfer)

Description

Initialize the non-negative factor loading with the label encoding by performing k-means on the centered and scaled data matrix concatenated. Then the comman factor matrix W is initilizated sequentially by nonnegative matrix fatorization using scaled but not centered data matrix.

Usage

initialize_params(X.list, r, W = NULL, verbose = TRUE)

Arguments

X.list	a list of ncells-by-ngenes gene expression matrix
r	scalar, dimensional of common factor matrix, which can be chosen as the rough number of identifiable cells types in the joint population (default 15).
W	ngenes-by-r numeric matrix. Supplied if parameter initialization is provided (default NULL).
verbose	boolean scalar, whether to show extensive program logs (default TRUE)

Value

a list containing initialized parameters: W, H.list, lambda.list, b.list

pengminshi/cFIT documentation built on July 11, 2021, 11:12 p.m.