CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC C C C ISOPAC Suite C C C C DR. STANLEY L. SCLOVE 312-996-2681 C C INFORMATION & DECISION SCIENCES DEPT. 312-996-2676 C C UNIVERSITY OF ILLINOIS AT CHICAGO C C 601 S. MORGAN ST. C C CHICAGO, IL 60607-7124 C C C C slsclove@uic.edu C http://www.uic.edu/classes/idsc/ids594/ISOPAC C C C C --------------------------------------------------------------- C C C TSPAC Library C http://www.uic.edu/classes/idsc/ids594/ISOPAC/TSPAC C C C PROGRAM TSSG1DTA C http://www.uic.edu/classes/idsc/ids594/ISOPAC/TSPAC/tssg1dta C C VERSION 1.7 2002: AUG 12 C VERSION 1.6 2002: AUG 12 C C VERSION 1.5 2002: AUG 10 C C VERSION 1.4 2002: JUL 30 C VERSION 1.3 2002: JUL 29 C C VERSION 1.2 2002: FEB 24 C VERSION 1.1 2002: JAN 4 C C VERSION 1.0 2001: DEC 28 C C C C PROGRAM TSSG1DTA WAS DEVELOPED FROM TSSGPDTA. C C C C THE "TSSGP" PROGRAMS ARE FOR SEGMENTING P-VARIATE C C TIME SERIES (NUMBER OF VARIABLES, P, GREATER THAN 1). FOR C UNIVARIATE TIME SERIES (P=1) THE "TSSG1" PROGRAMS ARE TO BE C C USED. C C TSPAC INCLUDES PROGRAMS USING DISTANCE IN THE METRIC OF THE C C ESTIMATED COMMON VARIANCE AS WELL AS PROGRAMS USING C C DIFFERENT VARIANCES, WITH ADJUSTMENT BY THE log VARIANCE C C I.E., USING ESTIMATED LOG LIKELIHOOD FOR THE C C GAUSSIAN MODEL WITH DIFFERENT VARIANCES. C C C C THERE ARE PROGRAMS IN MANUAL MODE, IN WHICH THE NUMBER OF C C CLASSES AND INITIAL MEANS ARE INPUT AND PROGRAMS IN AUTOMATIC C C MODE, WHICH TRY A RANGE OF NUMBERS OF CLASSES, WITH AUTOMATIC C C SETTING OF INITIAL MEANS. The output from the "A" series is C necessarily terse; for verbose output, use TSSG1DT for a C single value of the number of clusters K rather than C C a range of values of K. C C C ------------------------------------------------------------- C C C C PROGRAM RESTRICTIONS (CAN BE MODIFIED): C C C N -- SERIES LENGTH, AT MOST 19999; C C IP -- NUMBER OF VARIABLES, AT MOST 20; C C K -- NUMBER OF CLASSES, AT MOST 29; C C ITER -- MAXIMUM NUMBER OF ITERATIONS, 25. C C C C C C CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC C C C C DIMENSION X( 19999) DIMENSION SUM(29) DIMENSION D(29),ICLUS( 19999) DIMENSION IOTA( 19999),JOTA(29) DIMENSION TITLE(18) DIMENSION NG(29),XMEAN(29), XMNOL(29) DIMENSION FMT(18) DIMENSION SS(29),SSD(29),STDEV(29),VAR(29) DIMENSION ICLSOL( 19999) DIMENSION ET(29) DIMENSION PG(29) C C00000000111111111122222222223333333333444444444455555555556666666666 C23456789012345678901234567890123456789012345678901234567890123456789 C C The array PG contains the precisions (reciprocal variances). C DIMENSION NT(29,29),IRSUM(29),TP(29,29) DIMENSION PROB(29) C C DOUBLE PRECISION WGSS,SSD,STDEV,VAR,SS,SUM,VARHAT C DOUBLE PRECISION P,DET,D,TEMP,F,CF C C DOUBLE PRECISION A, ET DOUBLE PRECISION PG DOUBLE PRECISION TP,PROB C CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC C C C C C CONTROL STATEMENTS: C C C C DATASET TITLE C C IVERB, in format (6X,I1) C C IVERB = 1 for verbose output, = 0 otherwise C C SERIES LENGTH, N, IN FORMAT (2X,I5) C C FMT, IN FORMAT (18A4), E.G., (1X,F4.1) C C "FMT" MAY ALSO BE USED FOR OUTPUT: ALLOW AT LEAST C C ONE BLANK AT THE BEGINNING FOR CARRIAGE CONTROL. C C DATA, BY TIME POINT, IN FORMAT SPECIFIED BY FMT C C DATA IS INDEXED BY TIME POINT. C C C C KL, MIN NUMBER OF CLASSES, IN FORMAT (3X,I2) C C KU, MAX NUMBER OF CLASSES, IN FORMAT (3X,I2) C C C CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC C C READ (5,28000) TITLE READ (5,16500) IVERB C C WRITE PROGRAM INFORMATION. WRITE (6,17000) WRITE (6,53000) WRITE(6,35000) TITLE IF (IVERB .EQ. 1) WRITE (6,16600) IF (IVERB .EQ. 0) WRITE (6,16700) C C READ SERIES LENGTH, N. READ(5,16000) N C WRITE(6,33000) N C C C READ DATA FORMAT. READ (5,28000) FMT WRITE (6,11000) FMT WRITE (6,20600) C C C READ DATA. C DO 300 I = 1,N READ (5,FMT) X(I) IF (I .EQ. 1) GO TO 100 GO TO 200 100 CONTINUE XMAX = X(1) XMIN = X(1) 200 CONTINUE IF (X(I) .LT. XMIN ) XMIN = X(I) IF (X(I) .GT. XMAX ) XMAX = X(I) 300 CONTINUE WRITE (6,54000) WRITE (6,FMT) XMIN WRITE (6,55000) WRITE (6,FMT) XMAX C C READ KL AND KU, THE MIN AND MAX NUMBER OF CLASSES TO BE TRIED. READ(5,15000) KL READ(5,15000) KU WRITE(6,18000) KL,KU C C SET CONSTANTS. C PI = 3.1415927 C C XN = N DO 400 INTEG=1,N IOTA(INTEG) = INTEG 400 CONTINUE C C --------------------------------------------- C C K is the number of classes of segment. C The algoritm will run for K between KL and KU. C DO 6500 K = KL,KU C C K IS THE NUMBER OF CLASSES OF SEGMENT. C C COMPUTE INITIAL MEANS: C DO 500 IG=1,K XMEAN(IG) = XMIN + IG*(XMAX - XMIN )/(K+1) 500 CONTINUE C WRITE (6,19000) K C WRITE INITIAL MEANS WRITE (6,20400) C DO 600 IG = 1,K WRITE (6,20500) IG WRITE (6,FMT) XMEAN(IG) 600 CONTINUE C C SET CONSTANTS FOR THIS VALUE OF K. C DO 700 J = 1,K JOTA(J) = J 700 CONTINUE C C PARAMETERS FOR MODEL WITH COMMON VARIANCE ARE C K MEAN VECTORS AND ONE VARIANCE. C IP = 1 NOPARM = K*IP + IP*(IP+1)/2 C PARAMETERS FOR MODEL WITH DIFFERENT COVARIANCE MATRICES ARE C K MEAN VECTORS AND K COVARIANCE MATRICES. So the number C NPRMDF of parameters is C NPRMDF = K*IP + K*IP*(IP+1)/2 C C NON-DISTRIBUTIONAL PARAMETERS: C K-BY-K TRANSITION PROBABILITY MATRIX C GIVES K(K-1) FREE TRANSITION PROBABILITIES:-- C NOPARM = NOPARM + K*(K-1) NPRMDF = NPRMDF + K*(K-1) C C DO 800 IG=1,K VAR(IG) = 1.0 C FOR FIRST ITERATION, MARGINAL DISTRIBUTION OF LABELS, PROB, C IS TAKEN TO BE UNIFORM. C PROB(IG) = 1.0/K 800 CONTINUE C C ITER = 1 C C ITERATIONS BEGIN HERE C --------------------- C 900 CONTINUE C IF (ITER .EQ. 1) GO TO 1100 C Store values from preceding iteration. DO 950 IG = 1,K XMNOL(IG) = XMEAN(IG) 950 CONTINUE DO 1000 I = 1,N C ICLSOL(I) = ICLUS(I) 1000 CONTINUE C SAVE PREVIOUS VALUE OF -2 LOG MAX LIKELIHOOD: XMN20L = XMN2LL 1100 CONTINUE DO 1105 L = 1,K IF (ITER .GT. 1) GO TO 1110 PG(L) = 1.0 GO TO 1105 1110 A = SSD(L)/NG(L) VAR(L) = A STDEV(L) = SQRT( VAR(L) ) IF( IVERB .EQ. 1) WRITE (6,71000) L, STDEV(L) 1105 CONTINUE C C COMMENCE DISTANCE COMPUTATIONS. C DO 2400 I = 1,N C C INITIALIZE DISTANCES AT ZERO. DO 1200 L = 1,K D(L) = 0.0 1200 CONTINUE C FOR FIRST ITERATION, EUCLIDEAN DISTANCE IS USED BECAUSE C NO VARIANCE ESTIMATE IS YET AVAILABLE. AFTER THE FIRST C ITERATION, DISTANCE WILL BE TAKEN IN THE METRIC OF THE C GROUP VARIANCE. C C DO 1600 L=1,K TEMP = XMEAN(L) - X(I) D(L) = TEMP**2/VAR(L) C ( VAR(L) has been set to 1 for first iteration. ) C 1600 CONTINUE C DO 2200 L = 1,K C FOR FIRST ITERATION, CLASSIFICATION IS SIMPLY BY C MINIMUM DISTANCE. AFTER THAT, THE TRANSITION C PROBABILITIES AND VARIANCES ENTER: C IF (ITER .EQ. 1) GO TO 2200 C C UP TO NOW, D(L) IS (SQUARED) DISTANCE C IT MUST NOW BE MODIFIED TO PROBABILITY, C FOR MULTIPLICATION BY THE TRANSITION PROBABILITIES. C ADD LOG OF VARIANCE TO DISTANCE. C ET(L) IS LOG OF VARIANCE OF GROUP L: D(L) = D(L) + ET(L) ARG = -D(L)/2.0 IF (ARG .LT. -180.2) GO TO 1800 GO TO 1900 1800 D(L) = 0.0 GO TO 2000 1900 CONTINUE C C MOVE FROM LOG PROB SCALE TO PROB SCALE: D(L) = EXP(ARG) C D(L) IS NO LONGER A DISTANCE: C IT IS (PROPORTIONAL TO) THE BELONGING PROBABILITY 2000 CONTINUE C IF (I .EQ. 1) GO TO 2100 IG1 = ICLUS(I-1) C D(L) = -TP(IG1,L)*D(L) GO TO 2200 2100 CONTINUE C CLASSIFY FIRST OBSERVATION D(L) = -D(L)*PROB(L) 2200 CONTINUE C F = D(1) ICLUS(I) = 1 DO 2400 L = 2,K IF ( D(L) - F ) 2300,2400,2400 2300 F = D(L) ICLUS(I) = L 2400 CONTINUE C IF (IVERB .EQ. 0) GO TO 2505 C WRITE (6,10000) C WRITE (6,12000) (IOTA(J), J=1,N) C WRITE (G,12000) (ICLUS(I), I=1,N) C IF (N .LE. 200) WRITE (6,10000) C IF (N .LE. 200) WRITE (6,12000) (IOTA(J), J=1,N) C IF (N .LE. 200) WRITE (6,12000) (ICLUS(I),I=1,N) 2505 CONTINUE C DO 2600 IG = 1,K NG(IG) = 0 SUM(IG) = 0.0 SS(IG) = 0.0 SSD(IG) = 0.0 2600 CONTINUE DO 2700 I = 1,N C IGROUP = ICLUS(I) NG(IGROUP) = NG(IGROUP) + 1 SUM(IGROUP) = SUM(IGROUP) + X(I) SS(IGROUP) = SS(IGROUP) + X(I)*X(I) 2700 CONTINUE C IF ( IVERB .EQ. 1) WRITE (6,27000) (NG(IG),IG=1,K) C WGSS = 0.0 2800 CONTINUE C DO 3100 IG = 1,K C If some cluster has become empty, that cluster number will C be printed, and the program will skip ahead. IF (NG(IG) .EQ. 0) GO TO 2900 GO TO 3000 2900 CONTINUE WRITE (6,60000) IG GO TO 6600 3000 CONTINUE C C COMPUTE MEAN VECTORS XMEAN(IG) = SUM(IG)/NG(IG) C COMPUTE SUMS-OF-SQUARED DEVIATIONS CF = SUM(IG)*SUM(IG)/NG(IG) SSD(IG) = SS(IG) - CF VAR(IG) = SSD(IG)/NG(IG) STDEV(IG) = SQRT( VAR(IG) ) IF( IVERB .EQ. 1) WRITE (6,64000) IG,XMEAN(IG),STDEV(IG) 3100 CONTINUE C C POOL: DO 3200 IG = 1,K WGSS = WGSS + SSD(IG) 3200 CONTINUE C C COMPUTE VARHAT, MLE OF COMMON VARIANCE: VARHAT = WGSS/N 3300 CONTINUE C XLGVAR = DLOG(VARHAT) XMN2LL = N*(IP*ALOG(2.0*PI) + IP + XLGVAR) C C IF (ITER .EQ. 1) GO TO 3600 C C C C DO 3500 I = 1,N C C IF (ICLUS(I) .EQ. ICLSOL(I)) GO TO 3500 C GO TO 3600 C3500 CONTINUE C If the clustering hasn't changed, GO TO 6000, C write the output for this value of K, and go on C to the next value of K. C C GO TO 6000 3600 CONTINUE C C IF NEW SEGMENTATION DIFFERS FROM OLD, C COMPUTE AND WRITE NEW STATISTICS C DO 3700 IG = 1,K IF (IVERB .EQ. 1) WRITE (6,26000) IG,XMEAN(IG) 3700 CONTINUE C C COMPUTE TRANSITION PROBABILITY MATRIX: C DO 3800 I1 = 1,K C DO 3800 J = 1,K NT(I1,J) = 0 3800 CONTINUE DO 3900 I = 2,N C IM1 = I-1 C IBEFOR = ICLUS(IM1) C IY = ICLUS(I) NT(IBEFOR,IY) = NT(IBEFOR,IY) + 1 3900 CONTINUE DO 4000 I1 = 1,K C IRSUM(I1) = 0 4000 CONTINUE DO 4100 I1 = 1,K C DO 4100 J = 1,K IRSUM(I1) = IRSUM(I1) + NT(I1,J) 4100 CONTINUE DO 4400 I1=1,K C XDENOM=IRSUM(I1) IF (XDENOM .EQ. 0.0) GO TO 4200 GO TO 4300 4200 XDENOM = K 4300 CONTINUE C DO 4400 J = 1,K XNUM = NT(I1,J) C IF THERE ARE NO TRANSITIONS FROM I1) THEN TP(I1,J) C IS SET EQUAL TO ZERO, FOR ALL J = 1,2,...,K. C 4400 TP(I1,J) = XNUM/XDENOM 4500 CONTINUE C C COMPUTE MARGINAL DISTRIBUTION OF LABELS: DO 4600 IG = 1,K PROB(IG) = NG(IG) PROB(IG) = PROB(IG)/N 4600 CONTINUE C TRANS = 0.0 DO 4700 I1=1,K C DO 4700 J=1,K ITEST = NT(I1,J) IF (ITEST .EQ. 0) GO TO 4700 IF (TP(I1,J) .EQ. 0.) GO TO 4700 TRANS = TRANS + NT(I1,J)*DLOG(TP(I1,J)) 4700 CONTINUE C TRANS = -2.0*TRANS C C If output is verbose, WRITE TRANSITION PROBABILITIES: IF ( IVERB .EQ. 0) GO TO 4950 WRITE (6,13000) WRITE (6,37000) (JOTA(JAY),JAY=1,K) DO 4800 I1 = 1,K C WRITE (6,38000) I1, (NT(I1,J),J=1,K) 4800 CONTINUE WRITE (6,39000) WRITE (6,37000) (JOTA(JAY),JAY=1,K) DO 4900 I1=1,K C WRITE (6,40000) I1, (TP(I1,J),J=1,K) 4900 CONTINUE WRITE (6,63000) (PROB(IG),IG=1,K) WRITE (6,32000) TRANS WRITE (6,49500) WRITE (6,50000) WRITE (6,22000) XMN2LL 4950 CONTINUE C C ACCOUNT FOR LABEL OF FIRST OBSERVATION: C C LABEL1 = ICLUS(1) PROBAB = PROB(LABEL1) FIRST = ALOG(PROBAB) C FIRST = -2.0*FIRST C C C COMPUTE MODEL SELECTION CRITERIA: C COMPUTE VALUES CORRESPONDING TO NEW SEGMENTATION AND OLD C VARIANCE (OMIT ON FIRST ITERATION) C IF (ITER .EQ. 1) GO TO 5200 C C FOR MODEL WITH COMMON VARIANCE C (NOT OPTIMIZED IN THIS PROGRAM; HOWEVER, IT IS CLEAR C THAT ONE SHOULD USE THE COMMON-VARIANCE MODEL IF C THE VALUES OF THE MODEL SELECTION CRITERIA HERE FOR THAT MODEL C ARE LESS THAN THOSE FOR THE MODEL WITH DIFFERENT COVARIANCE C MATRICES): C AICOLD = XMN2OL + TRANS + FIRST + 2.0*NOPARM SCHOLD = XMN2OL + TRANS + FIRST + ALOG(XN)*NOPARM IF (IVERB .EQ. 1) WRITE (6,59000) NOPARM IF (IVERB .EQ. 1) WRITE (6,46000) AICOLD IF (IVERB .EQ. 1) WRITE (6,47000) SCHOLD 5200 CONTINUE AIC = XMN2LL + TRANS + FIRST + 2.0*NOPARM IF (IVERB .EQ. 1) WRITE (6,48000) AIC SCH = XMN2LL + TRANS + FIRST + ALOG(XN)*NOPARM IF (IVERB .EQ. 1) WRITE (6,49000) SCH C C C COMPUTE PRECISION FOR ALL GROUPS. C IP = 1 TERM=0.0 DO 5700 L=1,K C If NG(L) is less than or equal to IP, which is 1, C the precision matrix cannot be updated; C the one from the preceding iteration will be retained. IF (NG(L) .GT. IP) GO TO 5300 WRITE (6,52000) L,NG(L) GO TO 5600 5300 CONTINUE C C THE VARIABLE A IS TEMP FOR VARIANCE OF GROUP L: A = SSD(L)/NG(L) VAR(L) = A 5400 CONTINUE 5500 CONTINUE C ET(L) = DLOG(A) C 5600 CONTINUE C TERM = TERM + NG(L)*ET(L) 5700 CONTINUE C COMPUTE MODEL SELECTION CRITERIA WITH DIFFERENT COVARIANCE C MATRICES: C C PARAMETERS: C K MEAN VECTORS OF DIMENSION P AND K P-BY-P COVARIANCE MATRICES, C WHERE P IS THE NUMBER OF VARIABLES XM2LLD = N*IP*ALOG(2*PI) + N*IP + TERM XM2LLD = XM2LLD + TRANS + FIRST C NPRMDF is number of parms under model with different C variances. SCHD = XM2LLD + ALOG(XN)*NPRMDF AICD = XM2LLD + 2.0*NPRMDF IF (IVERB .EQ. 0) GO TO 5750 WRITE (6,23000) WRITE (6,59000) NPRMDF WRITE (6,24000) XM2LLD WRITE (6,57000) AICD WRITE (6,58000) SCHD 5750 CONTINUE C C IF (IVERB .EQ. 1) WRITE (6,25000) ITER DO 5800 IG = 1,K IF (IVERB .EQ. 1) WRITE (6,21000) IG, XMEAN(IG) 5800 CONTINUE ITER = ITER + 1 C If 50 iterations already, stop and say so. IF (ITER .GE. 50) GO TO 5900 IF ( ITER .LT. 50 ) GO TO 5825 5825 CONTINUE C UNLESS 50 ITERATIONS HAVE ALREADY BEEN PERFORMED, CHECK FOR C CONVERGENCE. C C If convergence has not yet occurred, GO BACK C TO STMT 900 AND DO ANOTHER ITERATION. C COMPARE NEW RESULTS WITH OLD: C First, compare new means with old by C computing the max relative change (up or down). TEST = ABS( XMNOL(1)/XMEAN(1) - 1.0 ) DO 3550 IG = 2,K TEMP = ABS(XMNOL(IG)/XMEAN(IG) - 1.0 ) IF ( TEMP .GT. TEST ) TEST = TEMP 3550 CONTINUE C "TEST" is greatest relative change in mean. C If "TEST" is larger than 0.01, go to stmt 900 C to do another iteration. IF ( TEST .GT. 0.01) GO TO 900 C If convergence, go to stmt 6025 and write results C for this value of K. IF ( TEST .LE. 0.01) GO TO 6025 C C GO TO 900 5900 WRITE (6,44000) C C C 6000 CONTINUE C You arrive at 6025 if there's been convergence. C You also arrive at 6025 if there hasn't been, C but that fact will have been indicated. C 6025 CONTINUE C C OUTPUT TO BE WRITTEN UPON CONVERGENCE FOR EACH VALUE OF K: C --------------------------------------------------------- C WILL ALSO BE WRITTEN IF NO CONVERGENCE, BUT THAT FACT WILL C HAVE BEEN INDICATED. C WRITE (6,41500) K WRITE (6,41000) ITER C Write class frequencies and means: DO 6050 IG = 1,K WRITE (6,74000) IG, NG(IG), XMEAN(IG) 6050 CONTINUE C C Write transition probabilities: WRITE (6,39000) DO 6075 IG = 1,K WRITE(6,40000) IG, (TP(IG,J),J=1,K) 6075 CONTINUE C Write marginal distribution of states: WRITE(6,63000) (PROB(IG),IG=1,K) WRITE (6,27000) (NG(IG),IG=1,K) C C For model with common variance: WRITE (6,49500) C WRITE (6,22500) WRITE (6,50000) WRITE (6,50500) NOPARM WRITE (6,48000) AIC WRITE (6,49000) SCH C C COMPUTE AND WRITE COMMON STD DEV: C WGMS = WGSS/(N-K) 6100 CONTINUE WGMS = SQRT(WGMS) WRITE (6,31500) WGMS 6200 CONTINUE C WRITE (6,62000) C For model with different variances: C COMPUTE AND WRITE CLASS STD DEVs: C DO 6400 L = 1,K WGMS = SSD(L)/(NG(L)-1) SD = SQRT(WGMS) 6300 CONTINUE WRITE (6,61000) L WRITE (6,31600) SD 6400 CONTINUE C WRITE (6,23000) WRITE (6,59000) NPRMDF WRITE (6,24000) XM2LLD WRITE (6,57000) AICD WRITE (6,58000) SCHD C C WRITE(6,66000) NM1=N-1 C DO 6700 I = 1,NM1 IT = I ITP1 = I+1 IF( ICLUS(IT) .EQ. ICLUS(ITP1) ) GO TO 6700 WRITE(6,65000) IT, ICLUS(IT), ICLUS(ITP1) 6700 CONTINUE C C WRITE (6,45000) K C C This concludes processing for this value of K. C Move to stmt number 6500 -- end of DO LOOP -- and C start new value of K. C 6500 CONTINUE C WRITE (6,42000) 6600 STOP C C ........................................................... C 10000 FORMAT(//1X,'SEGMENTATION:'/) 11000 FORMAT(/1X,'DATA READ IN UNDER FORMAT ',18A4) 12000 FORMAT(1X, (40I3/) ) 13000 FORMAT(//1X,'TRANSITIONS') 15000 FORMAT(3X,I2) 16000 FORMAT(2X,I5) 16500 FORMAT(6X,I1) 16600 FORMAT(' VERBOSE OUTPUT') 16700 FORMAT(' TERSE OUTPUT') 17000 FORMAT('1','-----------------------------------------------'///, X1X,'TSPAC LIBRARY IN ISOPAC SUITE '// X1X,'PROGRAM TSSG1DTA '/ X,1X,'FOR TIME SERIES SEGMENTATION '/ X,1X,'USING GAUSSIAN DISTRIBUTIONS WITH DIFFERENT VARIANCES'/ X,1X,' '/ X1X,'DEVELOPED AND PROGRAMMED BY '// X1X,' DR. STANLEY L. SCLOVE 312/996-2681'/ X1X,' IDS DEPT (MC 294) 312/996-2676'/ X1X,' UNIVERSITY OF ILLINOIS AT CHICAGO '/ X1X,' 601 S. MORGAN ST. '/ X1X,' CHICAGO, IL 60607-7124 '// X1X,'VERSION 1.6 2002: AUG 12 '/ X1X,'---------------------------------------------------------') 18000 FORMAT('1',1X,'MIN AND MAX NUMBER OF CLASSES TO BE TRIED ARE ', XI2,' AND ',I2/) 19000 FORMAT(1H1,'K = ',I2,' CLASSES'/) 20000 FORMAT(1X,'INITIAL MEAN FOR CLASS ',I2,': ',(4F11.1/)) 20400 FORMAT(' INITIAL MEANS ') 20500 FORMAT(1H0,' CLASS ', I2) 20600 FORMAT(' ABBREVIATIONS: '/ X1X,' LL: log max likelihood '/ X1X,' AIC: Akaike Information Criterion '/ X1X,' SIC: Schwarz Information Criterion (BIC)'/) 21000 FORMAT(1X,'MEAN FOR CLASS ',I2,': ',(8F9.3/)) 22000 FORMAT(' - 2 LL FOR MODEL WITH COMMON VARIANCE = ',F13.1) C 22500 FORMAT(' FOR MODEL WITH COMMON VARIANCE: ') 23000 FORMAT(1X,'FOR MODEL WITH DIFFERENT VARIANCES:') 24000 FORMAT(' - 2 LL FOR MODEL WITH DIFFERENT VARIANCES = ',F13.1) 25000 FORMAT(/,1X,'ITERATION ', I2,/) 26000 FORMAT(1X,'MEAN FOR CLASS ',I2,': ',(8F9.3/)) 27000 FORMAT(1X,'NUMBERS IN CLASSES:'/,(9I12/)/) 28000 FORMAT(18A4) 29000 FORMAT(3X,I2) 30000 FORMAT(1X,'COMMON VARIANCE (DIVISOR IS DF):') 31000 FORMAT(1X,(8F9.3/)) 31500 FORMAT(' ESTIMATE OF COMMON STD DEV ', F9.3) 31600 FORMAT(1X,(8F9.2/)) 32000 FORMAT(' CONTRIBUTION OF TRANS. PROBS. TO LOG LIKELIHOOD =', XF11.1) 33000 FORMAT(1X,'SERIES LENGTH = ',I5) 35000 FORMAT(18A4) 36000 FORMAT(9X,(9I7/)) 37000 FORMAT(5X,(9I7/)) 38000 FORMAT(1X, I4, (9I7/)) 39000 FORMAT(/,1X,'TRANSITION PROBABILITIES') 40000 FORMAT(1X, I4,3X,(9F7.4/)) C 41000 FORMAT(/1X,'CONVERGENCE: AFTER ', X'ITERATION ',I2,'. RESULTS ARE PRINTED BELOW.'/) C 41500 FORMAT(' RESULTS FOR ',I2,' CLASSES OF SEGMENT ') 42000 FORMAT(1X,'PROGRAM ENDED SUCCESSFULLY.') 44000 FORMAT(1X,'ROUTINE HAS NOT CONVERGED IN PRESCRIBED ', X'MAX NUMBER OF ITERATIONS. STOP') 45000 FORMAT(//,1X,'PROGRAM ENDED SUCCESSFULLY FOR THE CASE ', X'K = ',I2,'.'///) 46000 FORMAT(1X,'AIC WITH NEW LABELS AND OLD DISTRIBUTIONAL ', X'PARAMETERS: ', F11.1) 47000 FORMAT(1X,'SIC WITH NEW LABELS AND OLD ', X'DISTRIBUTIONAL PARAMETERS: ', F11.1) 48000 FORMAT(1X,'AIC WITH NEW LABELS AND NEW DISTRIBUTIONAL ', X'PARAMETERS: ',F11.1) 49000 FORMAT(1X,'SIC WITH NEW LABELS AND NEW ', X'DISTRIBUTIONAL PARAMETERS: ',F11.1) 49500 FORMAT(//) 50000 FORMAT(//1X,'FOR MODEL WITH COMMON VARIANCE ', X'(NOT OPTIMIZED IN THIS PROGRAM): ') 50500 FORMAT(1X,'NUMBER OF PARAMTERS = ', I5) 51000 FORMAT(1X,'AIC FOR MODEL WITH COMMON VARIANCE = ',F11.1) 52000 FORMAT(/1X,'CLASS ',I2,' CONTAINS ONLY ',I4,' OBSERVATIONS: ', X'PRECISION FROM PREVIOUS ITERATION IS BEING RETAINED') 53000 FORMAT(/1X,'THE MODEL WITH COMMON VARIANCE IS NOT'/ X1X,'OPTIMIZED IN THIS PROGRAM; HOWEVER, IT IS CLEAR'/ X1X,'THAT ONE SHOULD USE IT IF HERE THE VALUES OF THE MODEL-'/ X1X,'SELECTION CRITERIA FOR THAT MODEL ARE LESS THAN THOSE '/ X1X,'FOR THE MODEL WITH DIFFERENT VARIANCES,'/ X1X,'WHICH IS OPTIMIZED HERE.'/) C 54000 FORMAT(1H0,1X,'MINIMUM: ') 55000 FORMAT(1H0,1X,'MAXIMUM: ') C 56000 FORMAT(1X,'SIC FOR MODEL WITH COMMON VARIANCE = ', F15.1) 57000 FORMAT(1X,'AIC FOR MODEL WITH DIFFERENT VARIANCES = ', F15.1) 58000 FORMAT(1X,'SIC FOR MODEL WITH DIFFERENT VARIANCES = ', F15.1) 59000 FORMAT(/,1X,'NUMBER OF PARAMETERS = ',I4/) 60000 FORMAT(1X,'NO OBSERVATIONS IN GROUP ',I3,'. STOP') C 61000 FORMAT(1H0,1X,'STANDARD DEVIATION FOR CLASS ',I2) C 62000 FORMAT(/,1X,'CLASS STANDARD DEVIATIONS ', X' (DIVISORS ARE ONE LESS THAN NUMBER IN GROUP):') 62500 FORMAT(/,1X,' CLASS STD DEVs: ') 63000 FORMAT(' MARGINAL PROB.VECTOR:'/,(9F11.4/)) 64000 FORMAT(1X,'GROUP ', I3, ' MEAN ',F9.2,' STDEV ', F9.2) 65000 FORMAT(' AT TIME ',I5,' CHANGED FROM ',I2, ' TO ', I2) 66000 FORMAT(' SEQUENCE OF CLASS LABELS ') 70000 FORMAT(1X,'GROUP ', I3, 'PRECISION ',F9.2) 71000 FORMAT(' CLASS ',I2,' STD DEV ',F9.3) 74000 FORMAT(' CLASS ',I2,' FREQ ',I5, ' MEAN ', F9.3) END