********> bugfix.55
Author:   Bill Ross
Date:     11/9/95

Programs: Anal

Severity: Slight

Problem:  Dihedral energies can be sometimes very slightly negative when
          in fact they should be 0. E.g. from an HP:

              -1.494231099963272E-09
              -1.915995362367084E-09
              -9.964292446706935E-10
              -1.059775878164970E-09
              -1.611578703819560E-09

          Thus if one sets the criterion for printing dihedral energies,
          EMAX(3), to 0 in hopes of seeing a listing of all dihedrals,
          some will be missing.


Affects:  ENERGY option.

Cause:    Possibly machine arithmetic.


Fix:  Make the following change to anala.f
-------------------------------------------------------------------------
*** OLD anala.f
--- NEW anala.f
***************
*** 2040,2045
        IF(IATG.GT.LATG) IJG = INMAT(IATG)+LATG
        IF(IATG.LE.LATG) IJG = INMAT(LATG)+IATG
        ETMAT(IJG) = ETMAT(IJG)+E
        IF(E.LT.EPMAX) GO TO 100
        APC = AP*CONVV
        PHS = PHASE(IC)*CONVV

--- 2040,2046 -----
        IF(IATG.GT.LATG) IJG = INMAT(IATG)+LATG
        IF(IATG.LE.LATG) IJG = INMAT(LATG)+IATG
        ETMAT(IJG) = ETMAT(IJG)+E
+       if (e.lt.0.0d0) e = 0.0d0
        IF(E.LT.EPMAX) GO TO 100
        APC = AP*CONVV
        PHS = PHASE(IC)*CONVV
-------------------------------------------------------------------------

Temporary workarounds: Setting EMAX(3) to be slightly negative (-0.1)
                       is recommended if one wants to print all dihedrals.

--