gputools (version 0.1-1)

gpuQr: Estimate the QR decomposition for a matrix

Description

This function estimates the QR decomposition for a matrix using column pivoting and householder matricies. The work is done on a GPU.

Usage

gpuQr(a)

Arguments

a

a matrix of floating point numbers. This is the matrix that will be decomposed into Q and R factors.

Value

• a list consisting of the following elements: Q, R, pivot, rank. Q is a matrix representing the Q factor of input a. R is a matrix representing the upper triangular R factor of input a. The pivot element is a vector representing the final permutation of the columns of a after the QR decomposition function finishes decomposing a. The format of pivot is similar to the base R function 'qr'. The rank element is a single integer representing an estimation of the rank of input matrix 'a' based on the results of the QR decomposition. In some rare cases, this rank can be wildly different from the actual rank of the matrix a. It is only an estimation.

References

Bjorck, Ake (1996) Numerical methods for least squares problems. SIAM.

Examples

# get some random data of any shape at all
	x <- matrix(runif(1000), 50, 20)
	qr <- gpuQr(x)
	print(qr)