Trailing-Edge
-
PDP-10 Archives
-
decuslib10-02
-
43,50145/drtmi.ssp
There are 2 other files named drtmi.ssp in the archive. Click here to see a list.
C DRTM 10
C ..................................................................DRTM 20
C DRTM 30
C SUBROUTINE DRTMI DRTM 40
C DRTM 50
C PURPOSE DRTM 60
C TO SOLVE GENERAL NONLINEAR EQUATIONS OF THE FORM FCT(X)=0 DRTM 70
C BY MEANS OF MUELLER-S ITERATION METHOD. DRTM 80
C DRTM 90
C USAGE DRTM 100
C CALL DRTMI (X,F,FCT,XLI,XRI,EPS,IEND,IER) DRTM 110
C PARAMETER FCT REQUIRES AN EXTERNAL STATEMENT. DRTM 120
C DRTM 130
C DESCRIPTION OF PARAMETERS DRTM 140
C X - DOUBLE PRECISION RESULTANT ROOT OF EQUATION DRTM 150
C FCT(X)=0. DRTM 160
C F - DOUBLE PRECISION RESULTANT FUNCTION VALUE DRTM 170
C AT ROOT X. DRTM 180
C FCT - NAME OF THE EXTERNAL DOUBLE PRECISION FUNCTION DRTM 190
C SUBPROGRAM USED. DRTM 200
C XLI - DOUBLE PRECISION INPUT VALUE WHICH SPECIFIES THE DRTM 210
C INITIAL LEFT BOUND OF THE ROOT X. DRTM 220
C XRI - DOUBLE PRECISION INPUT VALUE WHICH SPECIFIES THE DRTM 230
C INITIAL RIGHT BOUND OF THE ROOT X. DRTM 240
C EPS - SINGLE PRECISION INPUT VALUE WHICH SPECIFIES THE DRTM 250
C UPPER BOUND OF THE ERROR OF RESULT X. DRTM 260
C IEND - MAXIMUM NUMBER OF ITERATION STEPS SPECIFIED. DRTM 270
C IER - RESULTANT ERROR PARAMETER CODED AS FOLLOWS DRTM 280
C IER=0 - NO ERROR, DRTM 290
C IER=1 - NO CONVERGENCE AFTER IEND ITERATION STEPS DRTM 300
C FOLLOWED BY IEND SUCCESSIVE STEPS OF DRTM 310
C BISECTION, DRTM 320
C IER=2 - BASIC ASSUMPTION FCT(XLI)*FCT(XRI) LESS DRTM 330
C THAN OR EQUAL TO ZERO IS NOT SATISFIED. DRTM 340
C DRTM 350
C REMARKS DRTM 360
C THE PROCEDURE ASSUMES THAT FUNCTION VALUES AT INITIAL DRTM 370
C BOUNDS XLI AND XRI HAVE NOT THE SAME SIGN. IF THIS BASIC DRTM 380
C ASSUMPTION IS NOT SATISFIED BY INPUT VALUES XLI AND XRI, THEDRTM 390
C PROCEDURE IS BYPASSED AND GIVES THE ERROR MESSAGE IER=2. DRTM 400
C DRTM 410
C SUBROUTINES AND FUNCTION SUBPROGRAMS REQUIRED DRTM 420
C THE EXTERNAL DOUBLE PRECISION FUNCTION SUBPROGRAM FCT(X) DRTM 430
C MUST BE FURNISHED BY THE USER. DRTM 440
C DRTM 450
C METHOD DRTM 460
C SOLUTION OF EQUATION FCT(X)=0 IS DONE BY MEANS OF MUELLER-S DRTM 470
C ITERATION METHOD OF SUCCESSIVE BISECTIONS AND INVERSE DRTM 480
C PARABOLIC INTERPOLATION, WHICH STARTS AT THE INITIAL BOUNDS DRTM 490
C XLI AND XRI. CONVERGENCE IS QUADRATIC IF THE DERIVATIVE OF DRTM 500
C FCT(X) AT ROOT X IS NOT EQUAL TO ZERO. ONE ITERATION STEP DRTM 510
C REQUIRES TWO EVALUATIONS OF FCT(X). FOR TEST ON SATISFACTORYDRTM 520
C ACCURACY SEE FORMULAE (3,4) OF MATHEMATICAL DESCRIPTION. DRTM 530
C FOR REFERENCE, SEE G. K. KRISTIANSEN, ZERO OF ARBITRARY DRTM 540
C FUNCTION, BIT, VOL. 3 (1963), PP.205-206. DRTM 550
C DRTM 560
C ..................................................................DRTM 570
C DRTM 580
SUBROUTINE DRTMI(X,F,FCT,XLI,XRI,EPS,IEND,IER) DRTM 590
C DRTM 600
C DRTM 610
DOUBLE PRECISION X,F,FCT,XLI,XRI,XL,XR,FL,FR,TOL,TOLF,A,DX,XM,FM DRTM 620
C DRTM 630
C PREPARE ITERATION DRTM 640
IER=0 DRTM 650
XL=XLI DRTM 660
XR=XRI DRTM 670
X=XL DRTM 680
TOL=X DRTM 690
F=FCT(TOL) DRTM 700
IF(F)1,16,1 DRTM 710
1 FL=F DRTM 720
X=XR DRTM 730
TOL=X DRTM 740
F=FCT(TOL) DRTM 750
IF(F)2,16,2 DRTM 760
2 FR=F DRTM 770
IF(DSIGN(1.D0,FL)+DSIGN(1.D0,FR))25,3,25 DRTM 780
C DRTM 790
C BASIC ASSUMPTION FL*FR LESS THAN 0 IS SATISFIED. DRTM 800
C GENERATE TOLERANCE FOR FUNCTION VALUES. DRTM 810
3 I=0 DRTM 820
TOLF=100.*EPS DRTM 830
C DRTM 840
C DRTM 850
C START ITERATION LOOP DRTM 860
4 I=I+1 DRTM 870
C DRTM 880
C START BISECTION LOOP DRTM 890
DO 13 K=1,IEND DRTM 900
X=.5D0*(XL+XR) DRTM 910
TOL=X DRTM 920
F=FCT(TOL) DRTM 930
IF(F)5,16,5 DRTM 940
5 IF(DSIGN(1.D0,F)+DSIGN(1.D0,FR))7,6,7 DRTM 950
C DRTM 960
C INTERCHANGE XL AND XR IN ORDER TO GET THE SAME SIGN IN F AND FR DRTM 970
6 TOL=XL DRTM 980
XL=XR DRTM 990
XR=TOL DRTM1000
TOL=FL DRTM1010
FL=FR DRTM1020
FR=TOL DRTM1030
7 TOL=F-FL DRTM1040
A=F*TOL DRTM1050
A=A+A DRTM1060
IF(A-FR*(FR-FL))8,9,9 DRTM1070
8 IF(I-IEND)17,17,9 DRTM1080
9 XR=X DRTM1090
FR=F DRTM1100
C DRTM1110
C TEST ON SATISFACTORY ACCURACY IN BISECTION LOOP DRTM1120
TOL=EPS DRTM1130
A=DABS(XR) DRTM1140
IF(A-1.D0)11,11,10 DRTM1150
10 TOL=TOL*A DRTM1160
11 IF(DABS(XR-XL)-TOL)12,12,13 DRTM1170
12 IF(DABS(FR-FL)-TOLF)14,14,13 DRTM1180
13 CONTINUE DRTM1190
C END OF BISECTION LOOP DRTM1200
C DRTM1210
C NO CONVERGENCE AFTER IEND ITERATION STEPS FOLLOWED BY IEND DRTM1220
C SUCCESSIVE STEPS OF BISECTION OR STEADILY INCREASING FUNCTION DRTM1230
C VALUES AT RIGHT BOUNDS. ERROR RETURN. DRTM1240
IER=1 DRTM1250
14 IF(DABS(FR)-DABS(FL))16,16,15 DRTM1260
15 X=XL DRTM1270
F=FL DRTM1280
16 RETURN DRTM1290
C DRTM1300
C COMPUTATION OF ITERATED X-VALUE BY INVERSE PARABOLIC INTERPOLATIONDRTM1310
17 A=FR-F DRTM1320
DX=(X-XL)*FL*(1.D0+F*(A-TOL)/(A*(FR-FL)))/TOL DRTM1330
XM=X DRTM1340
FM=F DRTM1350
X=XL-DX DRTM1360
TOL=X DRTM1370
F=FCT(TOL) DRTM1380
IF(F)18,16,18 DRTM1390
C DRTM1400
C TEST ON SATISFACTORY ACCURACY IN ITERATION LOOP DRTM1410
18 TOL=EPS DRTM1420
A=DABS(X) DRTM1430
IF(A-1.D0)20,20,19 DRTM1440
19 TOL=TOL*A DRTM1450
20 IF(DABS(DX)-TOL)21,21,22 DRTM1460
21 IF(DABS(F)-TOLF)16,16,22 DRTM1470
C DRTM1480
C PREPARATION OF NEXT BISECTION LOOP DRTM1490
22 IF(DSIGN(1.D0,F)+DSIGN(1.D0,FL))24,23,24 DRTM1500
23 XR=X DRTM1510
FR=F DRTM1520
GO TO 4 DRTM1530
24 XL=X DRTM1540
FL=F DRTM1550
XR=XM DRTM1560
FR=FM DRTM1570
GO TO 4 DRTM1580
C END OF ITERATION LOOP DRTM1590
C DRTM1600
C DRTM1610
C ERROR RETURN IN CASE OF WRONG INPUT DATA DRTM1620
25 IER=2 DRTM1630
RETURN DRTM1640
END DRTM1650