build_lassos: Calculate 2D fused-lasso solutionpath for a countmatrix and...

In tmirus/TTT: What the Package Does (One Line, Title Case)

Description

Calculate 2D fused-lasso solutionpath for a countmatrix and save as object This function was adapted with some changes from Martin Fahrenberger's sflST package (https://github.com/Martin-Fahrenberger/sflST)

Usage

build_lassos(counts, ids, name, output_folder = NULL, ncores = 4, gamma = 1)

Arguments

counts	input matrix of counts with genes in columns and spot identifier in rows
ids	coordinate table which converts the spot identifier into an x and y coordinate on a grid
name	Name of the experiment
output_folder	folder in which to save the results
ncores	number of parallel processes to start (default 8)
gamma	ratio of penalization between sparsity and continuity (default 1, high = sparse)

Details

This function builds a solution path for each gene (column) in a input matrix, selects the models according to the BIC and builds a count matrix

Value

list containing three elements

1. counts - count matrix in the same format as input count matrix
   

2. ids - barcode data frame that assigns spot identifiers to locations on the grid

3. lls - numeric vector containing least squares fit of model to raw expression for each gene

tmirus/TTT documentation built on April 17, 2021, 11:04 p.m.