SUBROUTINE CAFOUT(WHERE)
C
C       J.T. Armstrong    29 Dec 1990
C
C       This prints the results from program BINFIT to unit INMOD
C       (e.g., a model with comments), and a summary to unit OUTC. 
C       Adapted for use with binary data from the Mt. Wilson
C       interferometer from BAFOUT by R.S. Simon
C
C-----------------------------------------------------------------------

      INCLUDE 'BINFIT.INC'

      CHARACTER*64 MODNAM, OLDMOD
      CHARACTER*24 MESSGE
      CHARACTER*80 HISTRY
      REAL*8 STAR
      INTEGER*4 VLBOPE, GETIN, IER, I, I1, I2, K, L, LEN1, IFORM, WHERE
      DATA STAR /'*   '/

C WHERE=1 => Write summary to terminal

      IF (WHERE.EQ.1) GO TO 200

C Otherwise, write to output model file:
C
C Close the input model file.
C
      INQUIRE (UNIT=INMOD, NAME=MODNAM)
      CLOSE (UNIT=INMOD,STATUS= 'DELETE')
      OLDMOD = MODNAM
C
C Open the output model file.
C
  17  CONTINUE 
      IER = VLBOPE(INMOD, MODNAM, 'MODEL', 'NEW', MODNAM)
      IER = 1
      IF (IER.NE.1) THEN
          IER = GETIN(MODNAM, ' Name for output model file: ', L)
          IF (IER.NE.1) CALL EXIT
          GOTO 17
      END IF
C
C Write the header in the output model file.
C
      CALL MKHIST('BINFIT', '  8 Jan 91', HISTRY) ! 1st 2 args <= 16 char
      L = LEN1(HISTRY)
      WRITE (INMOD,1000) '! ',HISTRY(1:L)
      L = LEN1(INDSN)
      WRITE (INMOD,1000) '! Input data file:   ',INDSN(1:L)
      L = LEN1(OLDMOD)
      WRITE (INMOD,1000) '! Input model file:  ',OLDMOD(1:L)
      WRITE (INMOD,1001) STARNAME,DATE
      WRITE (INMOD,1000) '!'
      WRITE (INMOD,1000) '!  Filter    Width     Reduced   Degrees of'
      WRITE (INMOD,1000) '!                    Chi squared   Freedom'
C
      DO IG=1,NFILT
          WRITE (INMOD,1002) LAMBDA0(IG),DLAMBDA(IG),FCHISQ(IG),
     1                       NZVIS(IG)-NVPAR(IG)
      END DO

      WRITE (INMOD,1003) ' Reduced chi squared for all data:  ',TCHISQ
      WRITE (INMOD,1004) ' Number of ind. data points:  ',NZTOT
      WRITE (INMOD,1004) ' Degrees of freedom:  ',NZTOT-NVPARTOT
      WRITE (INMOD,1000) '!'
      WRITE (INMOD,1000) '! Starting Model (v => variable parameter):'
      CALL CWRITEMD(INMOD,'!',PARMSTRT,0 0)
      WRITE (INMOD,1000) '!'
      IF (CONF.GT..5) THEN
          WRITE(INMOD,1005) CONF*100.
          WRITE(INMOD,1000) '! estimated from covariance matrix: '
          CALL CWRITEMD(INMOD,'!',PUNCERT0,0,0)
      END IF
      WRITE (INMOD,1000) '! Final Model:'
C Write the final model.
      CALL CWRITEMD(INMOD,' ',PARM,0,0)

  200 CONTINUE                    ! Enter here for summary to terminal
      DO IC = 1, NCOMP
          DO IG = 1, NFILT
              DO K=1,8
                  IF (VARY(IC,IG,K)) THEN
                      FLAG(IC,IG,K) = VARIED
                  ELSE
                      FLAG(IC,IG,K) = FIXED
                  ENDIF
              END DO
          END DO
      END DO
      WRITE (OUTC,1014) STARNAME, ITER, METHOD(1:19)
      CALL CWRITEMD(OUTC,' ',PARM,0,0)

C Generate comment on quality of fit:
      IF (TCHISQ .LT. 0.5) THEN
          MESSGE = 'Too good to be true!'
      ELSEIF (TCHISQ .LT. 1.) then
          MESSGE = 'Astoundingly good!'
      ELSEIF (TCHISQ .LT. 2.) THEN
          MESSGE = 'I''m impressed.'
      ELSEIF (TCHISQ .LT. 3.) THEN
          MESSGE = 'Good'
      ELSE
          MESSGE = ' '
      ENDIF
C
      WRITE (OUTC,1003) 'Total reduced chi squared:  ',TCHISQ
C

      RETURN
C
 1000 FORMAT (A,A,A,A)
 1001 FORMAT ('! Source: ',A,';   Date: ',I6)
 1002 FORMAT ('!',3X,F5.1,' nm',2X,F3.0,' nm',3X,F8.3,3X,I4)
 1003 FORMAT ('!',(A), F8.3 )
 1004 FORMAT ('!',(A), I4 )
 1005 FORMAT ('! Levenberg-Marquardt uncertainties at',F5.1,
     1         ' percent confidence level,')
 1014 FORMAT (80('-')/' Source: ',A,'  Iteration #',I3,6X,A)
      END