Matrix Inversion, Real Numbers

8-72

A. E. Sapega, Trinity College, Hartford, Connecticut

The program inverts a matrix, up to size 12 x 12, of real

numbers. The algorithm used is the Gauss-Jordan method.

A unit vector of appropriate size is generated internally at

each stage. Following the Gauss sweep-out, the matrix is

shifted in storage, another unit vector is generated and the

calculation proceeds.

Minimum Hardware:
PDP-8

Other Programs Needed:
FORTRAN Compiler and FORTRAN

Operating System

Storage Requirement:
Uses all of core not used by the

FORTRAN Operating System.

Execution Time:
Actual computation takes less than

10 seconds. Data read-in and

lead-out may take up to five minutes.

Catalog: November 1969

Minimum Hardware:	PDP-8
Other Programs Needed:	FORTRAN Compiler and FORTRAN Operating System
Storage Requirement:	Uses all of core not used by the FORTRAN Operating System.
Execution Time:	Actual computation takes less than 10 seconds. Data read-in and lead-out may take up to five minutes.
Catalog:	November 1969