Even though spelling the name correctly (Andersen vs. Anderson) does not improve your understanding about the subject, it does somewhat show how careful you are about the subject.
If you are interested in equilibrium NVT properties, the AndersEn thermostat is often fine and correct. If you are interested in dynamical properties, I will argue that you should not use a thermostat at all (even with the ones that give smooth deterministic trajectories).
Even though spelling the name correctly (Andersen vs. Anderson) does not
improve your understanding about the subject, it does somewhat show how
careful you are about the subject.
This mini-attack feels unjustified. Hope your Russian is better than his
English. A name is just a tag given to us by our parents.
If you are interested in equilibrium NVT properties, the AndersEn
thermostat is often fine and correct. If you are interested in dynamical
properties, I will argue that you should not use a thermostat at all (even
with the ones that give smooth deterministic trajectories).
Why not just writing the above explanation?
Carlos
many thanks to everyone for the useful remarks.
I have another relevant to the topic of this thread question.
When I draw new velocities every say 100 MD steps via
run 2000 every 100 “velocity all create 235.0 4928459 rot yes dist gaussian”
is the random number generator initialized only once with the seed 4928459 when the “velocity all create …” is invoked for the first time, or it is initialized every 100 steps anew with the same seed so that I get the same distribution of velocities drawn every 100 steps?
Thank you in advance!
Anton
many thanks to everyone for the useful remarks.
I have another relevant to the topic of this thread question.
When I draw new velocities every say 100 MD steps viarun 2000 every 100 "velocity all create 235.0 4928459 rot yes dist
gaussian"is the random number generator initialized only once with the seed 4928459
when the "velocity all create ..." is invoked for the first time, or it is
initialized every 100 steps anew with the same seed so that I get the same
distribution of velocities drawn every 100 steps?
it is always re-initialized with the same seed.