Google
 

Trailing-Edge - PDP-10 Archives - decuslib20-02 - decus/20-0026/convt.doc
There are 2 other files named convt.doc in the archive. Click here to see a list.
SUBROUTINE CONVT

PURPOSE
   CONVERT NUMBERS FROM SINGLE PRECISION TO DOUBLE PRECISION
   OR FROM DOUBLE PRECISION TO SINGLE PRECISION.

USAGE
   CALL CONVT (N,M,MODE,S,D,MS)

DESCRIPTION OF PARAMETERS
   N	- NUMBER OF ROWS IN MATRICES S AND D.
   M	- NUMBER OF COLUMNS IN MATRICES S AND D.
   MODE - CODE INDICATING TYPE OF CONVERSION
	    1 - FROM SINGLE PRECISION TO DOUBLE PRECISION
	    2 - FROM DOUBLE PRECISION TO SINGLE PRECISION
   S	- IF MODE=1, THIS MATRIX CONTAINS SINGLE PRECISION
	  NUMBERS AS INPUT.  IF MODE=2, IT CONTAINS SINGLE
	  PRECISION NUMBERS AS OUTPUT.	THE SIZE OF MATRIX S
	  IS N BY M.
   D	- IF MODE=1, THIS MATRIX CONTAINS DOUBLE PRECISION
	  NUMBERS AS OUTPUT.  IF MODE=2, IT CONTAINS DOUBLE
	  PRECISION NUMBERS AS INPUT.  THE SIZE OF MATRIX D IS
	  N BY M.
   MS	- ONE DIGIT NUMBER FOR STORAGE MODE OF MATRIX
	    0 - GENERAL
	    1 - SYMMETRIC
	    2 - DIAGONAL

REMARKS
   MATRIX D CANNOT BE IN THE SAME LOCATION AS MATRIX S.
   MATRIX D MUST BE DEFINED BY A DOUBLE PRECISION STATEMENT IN
   THE CALLING PROGRAM.

SUBROUTINES AND FUNCTION SUBPROGRAMS REQUIRED
   NONE

METHOD
   ACCORDING TO THE TYPE OF CONVERSION INDICATED IN MODE, THIS
   SUBROUTINE COPIES NUMBERS FROM MATRIX S TO MATRIX D OR FROM
   MATRIX D TO MATRIX S.