Hi all,
I need to generate different waiting time following the distribution
Pw(τ) = λexp(−λτ)
for particle undergoing brownian simulation.
I input \lambda from the input script ( 0.1, 0.2 1.0 2.0 etc) and using random number uniform number (u) to create these waiting time
τ = −log(1 − u)/(λ∆t)
\Delta t is integration time step.
Is it correct way of generation these exponentially distributed time ? (unitless ?).
Best
Parvathy
It is irrelevant if somebody here says “yay” or “nay” to your question. It is still going to be your responsibility to confirm whether your input is correct or not.
So you have to apply the “scientific method™”:
construct a sufficiently small test case that would give the correct result if your hypothesis is correct and a significantly wrong result in case it is not; then run the test and check the results. If inconclusive, construct a different test and try again until you have convinced yourself that your choice is adequate.
Hi all,
Probably I wanted to ask a different question. For example, if I need to generate waiting time, with a lambda as the rate parameter. So how Lammps will read this number from input file, if I use LJ units ? Will it take as inverse of scaled time ? Is it already in scaled unit ?
Thank you
LAMMPS will use whatever units the documentation of the command says the keyword is using and and how those are defined for the choice of units that you select with the units command. Again, if you want to be certain, apply the scientific method.