Multiwfn official website: //www.umsyar.com/multiwfn. Multiwfn forum in Chinese: http://bbs.keinsci.com/wfn
You are not logged in.
Pages: 1
Thanks! I was about to suggest the common trick of adding some small quantity to the denominator:
fragsurvar=fragsurvar/(fragsurarea+0.000000000000001)
But your solution will definitely be more stable and less sensitive to numeric errors!
Hi again... I opened the code and found these two lines in surfana.f90 file:
fragsurvar=fragsurvar/fragsurarea
fragsurvar(:,1)=fragsurvar(:,2)+fragsurvar(:,3)
¿Could this be that "fragsurarea" is 0 for positive values (the surface area corresponding to positive values of the LEA is zero, since LEA is negative everywhere, so fragsurvar is NaN for positive values, and accordingly the first element of the vector is also NaN since it is calculated by the sume in the second instruction? Maybe for LEA the second instruction should be changed to:
fragsurvar(:,1)=fragsurvar(:,3)
or even an if..else could be used to determine wether to use one or the other alternative depending on NaN values in fragsurvar(:,2)...
analysing the LEA of each atom surface using QAMS (sequence of options: 12 2 4 0 11) I obtained results like this:
Note: Average and variance below are in eV and eV^2 respectively
Atom# All/Positive/Negative average All/Positive/Negative variance
3 -36.41248 NaN -36.41248 NaN NaN 24.17663
4 -36.25866 NaN -36.25866 NaN NaN 21.78491
5 -35.78661 NaN -35.78661 NaN NaN 23.29892
6 -35.81292 NaN -35.81292 NaN NaN 23.65038
12 -22.13457 NaN -22.13457 NaN NaN 8.27424
13 -21.98016 NaN -21.98016 NaN NaN 14.85794
While I find reasonable the results in the first 3 columns (basically, the "positive" values are all NaN and not used in calculating the average, so the average of all values and of negative values are identical), I wonder if the other 3 columns are correct. I would expect a similar result than with the others: NAN for "positive", and therefore identical variance considering all points or only negative points. Could this be a problem and the "All variance" column should also match the "Negative variance" column?
Thanks
Luis
I have been experimenting with different files and observed that the problem is with ORCA calculations involving at least d functions (6-31G is ok but 6-31G* not), and not with wavefunctions generated by other programs (with Gaussian16 the result is OK). I also noticed that although ORCA uses 5 d functionals, Multiwfn uses 6. There is no problem with this transformation, at least the tests in the appendix 5 of the manual are passed, but could it be related with a problem generating the .47 file when it is read by NBO5?
Thanks
Luis
Thanks for testing! I will make more tests and if I figure out the problem I will post the solution
Best
Luis
sorry for not including this before!
I also repeated the calculation with the fch file in the /examples/ folder, and it works well.... so it seems that it is not a problem with NBO but probably with orca_2mkl program...
Thanks
Luis
thanks a lot for the answer!
I am using NBO 5.0 (Is pretty old, but the one for which I have a license). The wavefunction that I am using comes from an ORCA 5.0 job, from which I used the gbw file to obtain a .molden.input file with orca_2mkl program (with -molden option). I opened this file with Multiwfn 3.8 -the version without the graphic interface- installed on centos Linux, next I used the sequence of options 100-2-8-(name of the 47 file)-1.
I tried different basis set (def2TZVPPD and def2SVP) with similar outcome.
Maybe the problem is on my NBO very old version, which may have a different definition of the precision of the data. Keyword "FORMAT=PRECISE", that is used to specify the format of the data in the file .47, is not recognized in such an old NBO version
Some of the lines of the 2nd order perturbation theory section:
SECOND ORDER PERTURBATION THEORY ANALYSIS OF FOCK MATRIX IN NBO BASIS
Threshold for printing: 0.50 kcal/mol
E(2) E(j)-E(i) F(i,j)
Donor NBO (i) Acceptor NBO (j) kcal/mol a.u. a.u.
===============================================================================
within unit 1
1. BD ( 1) O 1- C12 52. RY*( 2) O 1 1368.32 1055.78 33.965
1. BD ( 1) O 1- C12 55. RY*( 5) O 1 21237.89 6696.45 337.074
1. BD ( 1) O 1- C12 56. RY*( 6) O 1 24462.6868992.73 1161.212
1. BD ( 1) O 1- C12 57. RY*( 7) O 1 4105.20 1249.59 64.019
1. BD ( 1) O 1- C12 60. RY*(10) O 1 ********11860.31 1045.866
1. BD ( 1) O 1- C12 62. RY*(12) O 1 ******** 8637.36 1464.453
Best
Luis
Trying to export a .47 file for its use with NBO, I noticed that the "Second Order Perturbation Theory Analysis of Fock Matrix in NBO Basis" obtained after runing gennbo over this .47 file yields unrealistic results, with lots of overflows. Other parts of the NBO calculation seem to be OK, so I wonder if there is any change on the wavefunction that should be applied before exporting the .47 file (maybe some spherical or cartesian conversions) or if there is any problem with the format of the exported Fock matrix (maybe some units) that I will have to set in the NBO input file to prevent this problem.
Thanks in advanced:
Luis
My bad! It turned out that I have not exported Multiwfnpath while running multfwn.... after doing it the problem (that was not dependent of using an ORCA file, since it also appeared with any of the files in the examples folder) was over!
I was trying to calculate the atomic an multipole moments in the "Fuzzy atomic space analysis" from a calculation done in ORCA5, but I noticed that the number of electrons is very large (several thousands for a small molecule that actually only includes 10 -and so it is determined by mutwfn when I load the molecule), and consequently the dipole moments, quadrupoles, etc, are also very very large.
What I am doing wrong? It seems that other functions are correctly calculated by multwfn.
Thanks
Luis
Pages: 1