tenxiterate: Visit and 'reduce()' all values in a 10xGenomics hdf5 file.

In mtmorgan/TENxGenomics: Interface to 10x Genomics data

Description

This function visits all records in the hdf5 file, passing a 'chunk' containing a list of value, row index, and column index to a user-provided FUN(). The chunks represent one or more complete cells (columns); the number of cells is controled by yieldSize. FUN() manipulates the arguments (typically a data reduction) in an arbitary way, returning the result.

Usage

tenxiterate(X, FUN, ..., yieldSize = 10000)

Arguments

X	A TENxGenomics-class instance. Rows and columns of X cannot be duplicated.
FUN	A function of at least one argument VALUE, used to summarize successive chunks of the hdf5 file. VALUE is a list with elements ridx (row index), cidx (column index) and value (count value). See [BiocParallel::bpiterate()].
...	Additional arguments, passed to FUN().
yieldSize	numeric(1) Number of columns (cells) to process during each iteration.

Details

All records in the hdf5 file are visited, but only those correspond to rows or columns of X are reported. The implementation uses [BiocParallel::bpiterate()].

mtmorgan/TENxGenomics documentation built on May 23, 2019, 8:19 a.m.