GAMESS plotting: DENDIF for two sets and specific orbitals

[Budong Peng <budong -AatT- chem.columbia.edu>]

 Dear Netters;
 I was trying to plot our the electronic density difference of two molecules,
 and especially for one or several orbitals. I followed the instruction from
 dendif.man, e.g., 'dendif' a formamide-H2O complex, and 'dendif' the
 formamide only (with geometry identical to that in the complex), and
 put MINUS=1 in the first .ddf file, with
 TWO
 TWO
 for PSISQR and SUBDEN, but it didn't work. The manual is as the
 following:
 ******************************************************************
            -AatT- PSISQR    {TWO|ONE|<occ.no.>:}
           ----------------------------------------------------------
           This card gives the electron occupation numbers of all the
           orbitals occupied in the whole molecule.  The actual
           number of orbitals is read from the whole molecule's
           PLTORB file.  If all occupation numbers are 2.0 or 1.0, a
           single shorthand keyword may be used.
           ----------------------------------------------------------
            -AatT- SUBDEN:   [TWO|ONE|<occ.no>:]
           ----------------------------------------------------------
           *** Repeat this read site MINUS times.
           This is the occupation number of each natural orbital
           occupied in a fragment of the whole molecule.  The
           meanings are the same as  -AatT- PSISQR.
 ********************************************************************
 I tried various formats for these two groups when I don't have identical
 occupation, but failed to figure out the right one.
 ********************************************************************
               The remaining input is read from the various PLTORB
           disk files.  You must run PLTORB a total of MINUS+1 times.
           The whole molecule's data is provided on unit 20, and the
           various fragments are read in from units 21, 22, 23, ...,
           with their electron occupancies provided  -AatT- SUBDEN in the
           exact same order.
 *******************************************************************
 I followed this part, if it doesn't mean that I have support all the
 parts.
 What's more, if I wnat to plot several orbitals, say, mostly related
 with the original O 2p orbitals, can I do that?
 Any suggestions and advice are greatly appreciated.
 Merry Christmas to you all.
 Budong Peng