prepare_job_array: Divide the entire data into many chunks to...

In churchill-lab/scRATE: A Bayesian model selection approach to test zero inflation from scRNA-seq data

Description

Divide the entire data into many chunks to "run_model_comparison" on a cluster in parallel

Usage

prepare_job_array(loomfile, num_chunks, outdir, dryrun, covariates = c(),
  layer = NULL, nCores = NULL, seed = NULL, gene_start = NULL,
  gene_end = NULL, chunk_start = NULL, chunk_end = NULL)

Arguments

loomfile	Expression quantity file (loom format).
num_chunks	Number of chunks.
outdir	Name of the folder where results should be stored.
dryrun	TRUE if you only want to check the job submission commands.
covariates	Name of covariates (A col.attrs name in the input loomfile.)
layer	A layer name of the count to use in the loom file.
nCores	Number of cores to run stan fitting in parallel
seed	Seed number to reproduce randomized results
gene_start	Starting gene index to analyze.
gene_end	Ending gene index to analyze.
chunk_start	Starting chunk index to submit.
chunk_end	Ending chunk index to submit.

Value

... None is returned.

churchill-lab/scRATE documentation built on Aug. 12, 2020, 2:52 p.m.