Thank you so much!

Unfortunately, using 100.19, that is not possible to construct promolecular density of different dimers without changing .wfn files.

I used that script for generating dimers with promolecular densities for comparing their densities with SCF-optimized densities (they are at H2p_*.wfn files).

for wfn in H2p_*.wfn;
do
  rm H0.5.wfn_001_0300_0434
  coord=`grep "(CENTRE  1)" $wfn | sed "s/ \+/ /g" | cut -d" " -f8`
  coord2=`echo "${coord} * 2" | bc -l`
  ADD=`echo "1000 10 48 6 31 0,0,-${coord} 1 32 0,0,${coord2} 1 1 -1 100 -5 1 ${wfn} 0" | tr " " "\n" | ./gMultiwfn H0.5.wfn 2>/dev/null | grep "Final result:" | sed "s/ \+/ /g" | cut -d" " -f4`
  echo "${wfn};${ADD}"
done

Probably, in the general case, I need a reader of XYZ-files and atomwfn directory for constructing promolecular densities for given XYZ's.

Best regards,
Igor

I just updated Multiwfn on Multiwfn website. I significantly extended the subfunction 19 of main function 100, it is highly recommended to use this function to combine fragment wavefunctions to form a whole wavefunction.

The patch in #2 has been considered in this update. The problem you reported in #1 has also been fixed.

Best regards,

Tian

0002-Copy-fragments-and-MOs-while-dublication-is.patch.txt

It will produce repeated independent structures with independent molecular orbitals.

Best regards,
Igor

I understand how to fix that part of the MultiWFN code for being it to work correctly (at least as I want).

I found a mistake with variable usage: center_tmp is not declared but used. The correct variable is ncenter_tmp.
There is a needed patch (works both for ifort and gfortran):
0001-center_tmp-is-not-declared.patch.txt

Best regards,
Igor

I'm trying to construct promolecule density of dimers using atom's densities.

I'm running the following set of commands:

./Multiwfn H0.5.wfn
6
31
0,0,-0.47243153
1
32
0,0,.94486306
1
1
0

It produces new.wfn which is incorrect since when I'm trying to integrate it using:

./Multiwfn new.wfn
100
4
1

it leads to SegFault. It comes from `noGUI/sub.f90:2610`

It seems to me that modifying of wfn by that way does not produce the correct wfn.
At least, the number of electrons should be 2 times more (occupation number).

There is an initial wfn and new.wfn:
H0.5.wfn - initial
new.wfn - wfn produced by MultiWFN

Best regards,
Igor