Enter your search terms Submit search form
	Web pdp-10.trailing-edge.com

SAMPLE MAIN PROGRAM FOR MULTIPLE REGRESSION - REGRE

PURPOSE
   (1) READ THE PROBLEM PARAMETER CARD FOR A MULTIPLE REGRES-
   SION, (2) READ SUBSET SELECTION CARDS, (3) CALL THE SUB-
   ROUTINES TO CALCULATE MEANS, STANDARD DEVIATIONS, SIMPLE
   AND MULTIPLE CORRELATION COEFFICIENTS, REGRESSION COEFFI-
   CIENTS, T-VALUES, AND ANALYSIS OF VARIANCE FOR MULTIPLE
   REGRESSION, AND (4) PRINT THE RESULTS.

REMARKS
   THE NUMBER OF OBSERVATIONS, N, MUST BE GREATER THAN M+1,
   WHERE M IS THE NUMBER OF VARIABLES.	IF SUBSET SELECTION
   CARDS ARE NOT PRESENT, THE PROGRAM CAN NOT PERFORM MULTIPLE
   REGRESSION.
   AFTER RETURNING FROM SUBROUTINE MINV, THE VALUE OF DETER-
   MINANT (DET) IS TESTED TO CHECK WHETHER THE CORRELATION
   MATRIX IS SINGULAR.	IF DET IS COMPARED AGAINST A SMALL
   CONSTANT, THIS TEST MAY ALSO BE USED TO CHECK NEAR-
   SINGULARITY.

SUBROUTINES AND FUNCTION SUBPROGRAMS REQUIRED
   CORRE  (WHICH, IN TURN, CALLS THE SUBROUTINE NAMED DATA)
   ORDER
   MINV
   MULTR

METHOD
   REFER TO B. OSTLE, 'STATISTICS IN RESEARCH', THE IOWA STATE
   COLLEGE PRESS', 1954, CHAPTER 8.