Skip to contents

sparseTools

R-CMD-check

Overview

sparseTools is an R package that implements an S4 class for sparse numeric vectors, called sparse_numeric.

Its use is to provide a class for sparse numeric vectors along with methods to work witht hose vectors without ever converting them to a dense vector

Rather than storing every zero, the class stores only:

  • the non-zero values
  • their positions
  • the full length of the vector

How the Package Works

A sparse_numeric object stores:

  • value: numeric vector of non-zero entries
  • pos: integer vector of the corresponding positions
  • length: full length of the dense vector

the package implements:

Coercions

  • Convert numeric → sparse: as(x, "sparse_numeric")
  • Convert sparse → numeric: as(x, "numeric")
  • as.numeric(x) method

Arithmetic

  • sparse_add(x, y)
  • sparse_sub(x, y)
  • sparse_mult(x, y)
  • sparse_crossprod(x, y) (dot product)
  • Overloaded operators: +, -, *

Mathematical utilities

  • mean(x)
  • norm(x) (Euclidean norm)
  • standardize(x) (z-score scaling)

Installation

Install the package with:

# install.packages("devtools")
devtools::install_github("SahithiVarma07/sparseTools")

Example Usage 

library(sparseTools)

# Create a sparse vector from a regular numeric vector
x <- as(c(0, 3, 0, 5), "sparse_numeric")
x
as.numeric(x)

# Compute mean (includes zeros)
mean(x)

# Euclidean norm
norm(x)

# Standardize the vector (z-score)
s <- standardize(x)
as.numeric(s)

# Arithmetic
y <- as(c(1, 0, 2, 0), "sparse_numeric")

as.numeric(x + y)
as.numeric(x - y)
as.numeric(x * y)

# Dot product
sparse_crossprod(x, y)