New implementation (June 2024) of TSH in Quantics

General
1

Hi Graham

Would you have a recent quantics input for a tsh calculation? I am not sure I am understanding well which keywords changed in the last version and I am getting errors when trying to run the butatriene lvc model with the new quantics version.

I attach the input and op file I was using in case you want to take a look.

1 Like
2

Hi Sandra,

what errors are you getting? You certainly need to replace “coupling nadvec” by “nadvec linear”. I will update the file in the quantics inputs.

3

The error reads

Error in reader_go_to_keyword subroutine.
Keyword "$method" not found in file "dynamics.in".

I just changed the input file and I get the same error message. Maybe it wants to do direct dynamics and I need to specify somehow that I have analytic potentials?

Thank you for your help!

4

In each of the trajectory directories in but_tsh/zagreb_trj/traj.n there should be a file “dynamics.in” which should contain

$system
 nstate =    2
 istate =    2
$surfhop                                                                        
 fssh adiabatic                                                                 
 nadvec linear                                                                  
 seed -1                                                                        
$dynamics                                                                       
 tstep = 0.1                                                                    
 max_time =   120.0000
$method
 qlib quantics
$output
 noprint  7,10,11
$end

Your error says the $method section is not there. Which version of the code are you using? In the latest, this section is now written by the code driving the TSH calculations (interfaces/prop_zagreb.F90). The latest master branch works for me.

5

Hmmm I am using our ctsh branch that we updated the other day (98b485380e85b8d741883062ec06f5a62f38b0da) but I am running normal surface hopping with it. It tells me there is nothing new to pull.

Maybe we forgot something or was anything updated after our meeting?

The dynamics.in in the zagreb_trj trajectory folders contains

$system
 nstate =    2
 istate =    2
~

which is definitely missing half of the file. If you think it would be useful I would have time during the day for a quick meeting (or tomorrow!).

6

I can see that the merge has created some seemingly redundant code in prop_zagreb and this could be causing issues with creating and using the input files. I’ll have a look asap and let you know.

7

There was a but in the prop_zagreb routine. The found array was not big enough and overwrote the do_ct flag causing problems with the writing of the input. Try the latest on your branch.

8

Thank you very much for your help. Now both our branch and the master give the same answers. However, I do not seem to manage to run TSH on precomputed potentials. Do you know whether there is any new keyword to do so? The interface seems to try to run a QM calculation and gives the error " Error, qm_energy file not found after QM calculation."

9

Is this problem with you CT TSH? The ZagHop code works fine with pre-computed potentials - at least the but_tsh test runs. However, with the older code you are using for the CTMQC, you need to specify the “code” to be used. I.e. the ctsh_inter interface. This is the lines

$method
 qm /home/graham/quantics_ctmqc/bin/ctsh_inter ../..

Do you have these, or are they not being written to the dynamics.in file?