NaN problem with m-GGA (TPSS)


Just Got Here
Hi.

Hope someone can help me.

I send a job, in 6.8 version, and it finishes without problem when I use GGA functional (PBE) or hybrid (PBE0), but when i send the same job with a m-GGA (TPSS) it begans to print NaN in many steps of the optimization and do not finishes properly.


This is my input

start
memory total 2000 Mb
basis spherical ; * library aug-cc-pvdz ; end
charge 0
dft
 direct
mult 1
xc xtpss03 ctpss03
end
driver
 maxiter 100
end
geometry noautosym noautoz
C 0.02243109 1.85869639 -0.01727099
C 1.20577937 1.13311771 0.00248712
C 1.15973320 -0.26331665 0.01379360
C -0.03808364 -0.99852963 0.00639339
C -1.20132012 -0.21357023 0.01463135
C -1.18939035 1.18354038 0.00069837
O -0.06394131 -2.27349920 -0.01638336
H 0.04396732 2.89009623 -0.04505021
H 2.11439855 1.62239149 0.00878670
H 2.05182691 -0.78196124 0.02833456
H -2.11451078 -0.69398116 0.03189978
H -2.07690060 1.71017236 0.00376005
H 0.73785167 -2.75835189 -0.02106111
end
task dft optimize
task dft freq


I got things like this

                        DFT ENERGY GRADIENTS

   atom               coordinates                        gradient
x y z x y z
1 C 0.042389 3.512427 -0.032637 NaN NaN NaN
2 C 2.278593 2.141282 0.004700 NaN NaN NaN
3 C 2.191578 -0.497596 0.026066 NaN NaN NaN
4 C -0.071968 -1.886947 0.012082 NaN NaN NaN
5 C -2.270166 -0.403589 0.027649 NaN NaN NaN
6 C -2.247622 2.236567 0.001320 NaN NaN NaN
7 O -0.120832 -4.296291 -0.030960 NaN NaN NaN
8 H 0.083086 5.461490 -0.085133 NaN NaN NaN
9 H 3.995634 3.065875 0.016604 NaN NaN NaN
10 H 3.877391 -1.477692 0.053545 NaN NaN NaN
11 H -3.995846 -1.311434 0.060282 NaN NaN NaN
12 H -3.924773 3.231757 0.007105 NaN NaN NaN
13 H 1.394337 -5.212529 -0.039800 NaN NaN NaN

                ----------------------------------------
| Time | 1-e(secs) | 2-e(secs) |
----------------------------------------
| CPU | 0.05 | 63.51 |
----------------------------------------
| WALL | 0.05 | 63.59 |
----------------------------------------

@ Step Energy Delta E Gmax Grms Xrms Xmax Walltime
@ ---- ---------------- -------- -------- -------- -------- -------- --------
@ 0 -307.53544646 0.0D+00 0.00000 NaN 0.00000 0.00000 555.4
                                    ok                ok       ok




How can i solve this?

Thanks in advance.

Just Got Here
It does not works for me.

I do not have this problem when i calculate with any GGA or hybrid, the problem appears when i use any mGGA

Forum Vet
Quote:Jcpkato Apr 5th 10:17 pm
It does not works for me.

I do not have this problem when i calculate with any GGA or hybrid, the problem appears when i use any mGGA


Jcpkato, I have found and fixed a problem in ctpss.
https://github.com/nwchemgit/nwchem/issues/108

If you find a problem in another metaGGA, please post your input here.
Thanks, Edo

Just Got Here
Hi Edoapra.

I tried your correlation vertion of TPSS and the problem is still appearing. I?l add the out in your github post. Hope you can help me.


Forum >> NWChem's corner >> Running NWChem