Solve Linear Equations and Invert Matrices

solve

Description

Solves a non-singular system of linear equations a %*% x == b or, if b is omitted, inverts the matrix a. The function solve is generic.

Usage

solve(a, b, ...) 
## S3 method for class 'solve':
default(a, b, tol = .Machine$double.eps, ...) 
## S3 method for class 'solve':
qr(a, b, ...)

Arguments

a	a non-singular square numeric or complex matrix, or a QR decomposition of such a matrix (as made by the qr function).
b	a numeric or complex (if a is complex) vector, or a matrix giving the right hand side of the equation to be solved. a %*% x == b. NROW(b) must match NCOL(a); b can have any number of columns. Its default value is the identity matrix of order NCOL(a), meaning the result of solve is the inverse of a.
tol	a small positive numeric value. If the estimated reciprocal condition number of a is less than tol, then the matrix is considered singular, and an error is given.
...	additional arguments passed to other methods for solve but are ignored by these methods.

Details

If given two arguments, a and b, the function returns the solution x for the system of equations a %*% x == b.
If given only a non-singular matrix a, the function returns its inverse.

The default method uses the LAPACK function dgesv or, for complex matrices, zgesv, to solve the system using the LU decomposition. See http://www.netlib.org/lapack/ for more details.

Value

returns the inverse of a or, if b is present, it returns the solution x to the system of equations a %*% x = b. It has the same dimensions as b.

Differences between Spotfire Enterprise Runtime for R and Open-source R

The LINPACK argument, which is available in open-source R, is ignored in Spotfire Enterprise Runtime for R.
If b contains missing values (NAs), open-source R solve(qr(a), b) returns a vector of zeros; in Spotfire Enterprise Runtime for R an error is returned.