processByChunks: Process data column by chunks

In cTRAP: Identification of candidate causal perturbations from differential gene expression data

Description

Columns will be processed per chunk if argument data is a perturbationChanges object containing a file path instead of a data matrix. Otherwise, the data will be processed as a single chunk.

Usage

processByChunks(data, FUN, ..., chunkSize = 10000)

Arguments

data	Data matrix or perturbationChanges object
FUN	Function: function to run for each chunk
...	Arguments passed to FUN
chunkSize	Integer: number of columns to load on-demand (a higher value increases RAM usage, but decreases running time)

Details

For instance, loading a chunk of 10000 CMap pertubations requires ~1GB of RAM compared to loading the whole dataset.

Value

Results of running FUN

cTRAP documentation built on Nov. 8, 2020, 10:58 p.m.