How to accelerate pepper process

General forum for EasySpin: questions, how to's, etc.
Post Reply
alwens
Newbie
Posts: 2
Joined: Sun Oct 22, 2017 6:49 am

How to accelerate pepper process

Post by alwens »

Hi all,

I want to simulate epr spectrum by using orca2easyspin. Then first I calculated epr tensors by orca (.prop file), and using orca2easyspin generate further data. I got such information:
Sys =
S: 1
xyz: [29X3 double]
Charge: 0
g: [2.0022 2.0028 2.0037]
gFrame: [0.1510 1.1134 -1.8332]
Nucs: 'N,C,H,C,H,N,C,C,C,C,C,H,C,N,H,H,H,C,C,C,H,C,H,C,H,C,H,C,N'
A: [29X3 double]
AFrame: [29X3 double]
Q: [29X3 double]
QFrame: [29X3 double]

to accelerate such simulation, i choose perturb method in pepper
Sys.lwpp =1; %mT
Exp.mwFreq =9.8; %GHz
Exp.Range = [315 368]; %mT
Opt.Method= 'perturb2';
[x,y]= pepper(Sys,Exp,Opt);


while three days passed, this work still in process. So i want to know are there any method to accelerate pepper process? My computer is linux system and 32 core, 128GB memory.


Thanks,
Best regards,
Sa
katarkon
Local Expert
Posts: 182
Joined: Mon Jan 12, 2015 4:01 am

Re: How to accelerate pepper process

Post by katarkon »

I suggest that spin system with 29 nuclei is extremely large. Firstly, all of C nuclei may be droped. Secondly, other nuclei with small HFC constants may be excluded too. The functions nucspinrmv or nucspinkeep may help in this. At third, I don't shure that perturbation method is appropriate choise for systems with S=1. I suggest that hybrid method should be used (matrix treatment for electron spins and perturbation for nuclear ones).
alwens
Newbie
Posts: 2
Joined: Sun Oct 22, 2017 6:49 am

Re: How to accelerate pepper process

Post by alwens »

katarkon wrote:I suggest that spin system with 29 nuclei is extremely large. Firstly, all of C nuclei may be droped. Secondly, other nuclei with small HFC constants may be excluded too. The functions nucspinrmv or nucspinkeep may help in this. At third, I don't shure that perturbation method is appropriate choise for systems with S=1. I suggest that hybrid method should be used (matrix treatment for electron spins and perturbation for nuclear ones).
thank you.
nwili
User
Posts: 25
Joined: Mon Oct 16, 2017 6:27 am

Re: How to accelerate pepper process

Post by nwili »

I suggest that spin system with 29 nuclei is extremely large.
Yes, most likely, the computer can't even store the Hamiltonian in Matrix Form, as the dimension is (2^30)x(2^30) (if all spins are 1/2).
Stefan Stoll
EasySpin Creator
Posts: 1041
Joined: Mon Jul 21, 2014 10:11 pm
Location: University of Washington

Re: How to accelerate pepper process

Post by Stefan Stoll »

orca2easyspin[c] takes a hyperfine cutoff as second parameter. E.g. if you set it to 0.5, all nuclei with hyperfine couplings less than 0.5 MHz will be removed from the spin system.
Post Reply