C NDTR 10
C.......................................................................NDTR 20
C NDTR 30
C SUBROUTINE NDTR NDTR 40
C NDTR 50
C PURPOSE NDTR 60
C COMPUTES Y = P(X) = PROBABILITY THAT THE RANDOM VARIABLE U,NDTR 70
C DISTRIBUTED NORMALLY(0,1), IS LESS THAN OR EQUAL TO X. NDTR 80
C F(X), THE ORDINATE OF THE NORMAL DENSITY AT X, IS ALSO NDTR 90
C COMPUTED. NDTR 100
C NDTR 110
C USAGE NDTR 120
C CALL NDTR(X,P,D) NDTR 130
C NDTR 140
C DESCRIPTION OF PARAMETERS NDTR 150
C X--INPUT SCALAR FOR WHICH P(X) IS COMPUTED. NDTR 160
C P--OUTPUT PROBABILITY. NDTR 170
C D--OUTPUT DENSITY. NDTR 180
C NDTR 190
C REMARKS NDTR 200
C MAXIMUM ERROR IS 0.0000007. NDTR 210
C NDTR 220
C SUBROUTINES AND SUBPROGRAMS REQUIRED NDTR 230
C NONE NDTR 240
C NDTR 250
C METHOD NDTR 260
C BASED ON APPROXIMATIONS IN C. HASTINGS, APPROXIMATIONS FOR NDTR 270
C DIGITAL COMPUTERS, PRINCETON UNIV. PRESS, PRINCETON, N.J., NDTR 280
C 1955. SEE EQUATION 26.2.17, HANDBOOK OF MATHEMATICAL NDTR 290
C FUNCTIONS, ABRAMOWITZ AND STEGUN, DOVER PUBLICATIONS, INC., NDTR 300
C NEW YORK. NDTR 310
C NDTR 320
C.......................................................................NDTR 330
C NDTR 340
SUBROUTINE NDTR(X,P,D) NDTR 350
C NDTR 360
AX=ABS(X) NDTR 370
T=1.0/(1.0+.2316419*AX) NDTR 380
D=0.3989423*EXP(-X*X/2.0) NDTR 390
P = 1.0 - D*T*((((1.330274*T - 1.821256)*T + 1.781478)*T - NDTR 400
1 0.3565638)*T + 0.3193815) NDTR 410
IF(X)1,2,2 NDTR 420
1 P=1.0-P NDTR 430
2 RETURN NDTR 440
END NDTR 450