RREF Reduced Row Echelon Form of a Matrix

Section: Array Generation and Manipulations


Calculates the reduced row echelon form of a matrix using Gauss Jordan elimination with partial pivoting. The generic syntax for rref is
   R = rref(A)

A default tolerance of max(size(A))*eps*norm(A,inf) is used to detect negligible column elements. The second form of rref returns a vector k as well as R

  [R,k] = rref(A)

where k is a vector that correponds to the columns of A used as pivot columns. If you want to control the tolerance used to identify negligible elements, you can use the form

  [R,k] = rref(A, tolerance)

This implementation of rref is based on the one from the matcompat lib for octave. It is copyright Paul Kienzle, and distributed under the GNU GPL.