Multiwfn forum

axeljr7

Member

Registered: 2023-07-31

Posts: 18

population analysis with MBIS partitioning

Dear all,
for the computation of MBIS charges implemented in multiwfn, is there a way to obtain and save the final converged parameters of the gaussian terms (sigm_i and N_i values)?
That would be highly beneficial and useful.

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,830

Website

Re: population analysis with MBIS partitioning

The MBIS code in Multiwfn currently is in experimental status, and was directly contributed by other people (Frank Jensen). I intend to polish and parallelize this piece of code soon, at that time perhaps I will add an option to export the intermediate quantities. Currently you have to manually modify the source code to realize this pupose (please search "subroutine MBIS" in population.f90)

axeljr7

Member

Registered: 2023-07-31

Posts: 18

Re: population analysis with MBIS partitioning

that sounds really exciting. Thanks for that in advance. If might be of help, one can also get the python scripts of the method provided by the main developers at: https://github.com/theochem/denspart
Additionally, one interesting option for the new implementation in Multiwfn would also be also to include it in fuzzy space computations and specially computations of effective and free volumes and atomic polarizability (option 15).  That would be highly appreciated.

The MBIS code in Multiwfn currently is in experimental status, and was directly contributed by other people (Frank Jensen). I intend to polish and parallelize this piece of code soon, at that time perhaps I will add an option to export the intermediate quantities. Currently you have to manually modify the source code to realize this pupose (please search "subroutine MBIS" in population.f90)

axeljr7

Member

Registered: 2023-07-31

Posts: 18

Re: population analysis with MBIS partitioning

I just tried to modify the code but compilation from the source code raises some errors (as can be seen in the photo attached).
I would be so thankful for a hint to fix it.

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,830

Website

Re: population analysis with MBIS partitioning

I just tried to modify the code but compilation from the source code raises some errors (as can be seen in the photo attached).
I would be so thankful for a hint to fix it.

https://i.postimg.cc/xkyxJXTT/Screenshot-at-2023-08-12-15-36-13.png

I don't find any problem in compiling the latest source code package on Multiwfn website. From the screenshot, mod_2f2.mod cannot be found, however, after compiling ext/2F2.f90 as shown by the following lines in Makefile, mod_2f2.mod should then be produced in current folder. Please check if "ext" subfolder exists in current folder.

2F2.f90.o : ext/2F2.f90 util.o Bspline.o
    $(FC) $(OPT) -c ext/2F2.f90 -o 2F2.f90.o

axeljr7

Member

Registered: 2023-07-31

Posts: 18

Re: population analysis with MBIS partitioning

I just tried to modify the code but compilation from the source code raises some errors (as can be seen in the photo attached).
I would be so thankful for a hint to fix it.

https://i.postimg.cc/xkyxJXTT/Screenshot-at-2023-08-12-15-36-13.png

I don't find any problem in compiling the latest source code package on Multiwfn website. From the screenshot, mod_2f2.mod cannot be found, however, after compiling ext/2F2.f90 as shown by the following lines in Makefile, mod_2f2.mod should then be produced in current folder. Please check if "ext" subfolder exists in current folder.

2F2.f90.o : ext/2F2.f90 util.o Bspline.o
    $(FC) $(OPT) -c ext/2F2.f90 -o 2F2.f90.o

ext subdirectory exists and includes no2F2.c, 2F2.h, 2F2.f90, 2F2.c and no2F2.f90 (but not no2F2.h)
Can it be the reason of the problem?

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,830

Website

Re: population analysis with MBIS partitioning

There is no file named no2F2.h.

If you cannot find a way to solve it, you can simply delete definition of "subroutine orbfracderiv" and "function fracderiv" from function.f90, and remove "userfunc=fracderiv(x,y,z)" from this file. Also, remove "use mod_2F2, only: set_alpha_level" from Multiwfn.f90.

axeljr7

Member

Registered: 2023-07-31

Posts: 18

Re: population analysis with MBIS partitioning

There is no file named no2F2.h.

If you cannot find a way to solve it, you can simply delete definition of "subroutine orbfracderiv" and "function fracderiv" from function.f90, and remove "userfunc=fracderiv(x,y,z)" from this file. Also, remove "use mod_2F2, only: set_alpha_level" from Multiwfn.f90.

I tried it but unfortunately didn't fix the issue. I will try it on another machine and write here the outcome.
Thanks again.

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,830

Website

Re: population analysis with MBIS partitioning

Now sigma_i and N_i values can be directly outputted, see http://sobereva.com/wfnbbs/viewtopic.php?pid=3813