read_sparse_csv: Read sparse (numeric) CSVs

In Laurae2/Laurae: Advanced High Performance Data Science Toolbox for R

Description

This function allows you to big load sparse numeric CSVs. Loading in chunks allows to not explode the memory as when the data is imported into R, it is typically a dense matrix. Verbosity is automatic and cannot be removed. In case you need this function without verbosity, please compile the package after removing verbose messages.

Usage

read_sparse_csv(input, iterfeature, nfeatures = NA, colClasses = NA,
  RDS = NA, compress_RDS = TRUE, NA_sparse = FALSE)

Arguments

input	The input file name.
iterfeature	The amount of variables loaded per iteration. The smaller the longer it takes to load the whole dataset in its entireity.
nfeatures	The IDs of features to load. Defaults to NA which means loading all columns.
colClasses	The classes of the columns. Defaults to NA which means autoselection as numeric. Do not modify (keep default).
RDS	Whether to store in a RDS file of that name. Defaults to NA which means no RDS file. Otherwise, it takes RDS as filename.
compress_RDS	Whether to compress RDS file. Defaults to TRUE
NA_sparse	Whether sparsity is defined as NA. Defaults to FALSE

Value

The sparse matrix

Examples

#read_sparse_csv("train_numeric.csv", iterfeature = 100, IDs = c(1:500, 601:1000), colClasses = NA,
#RDS = TRUE, compress_RDS = FALSE, NA_sparse = FALSE)

Laurae2/Laurae documentation built on May 8, 2019, 7:59 p.m.