Thank you for your response.

> Thank you for your response.

> But I have a question, what do you mean by a very large N? how large?

without the finite size correction, the limit of g(r) for large r is not

1.0, but (n-1)/n. this is typically a small deviation.

I repeated the simulation for a large box but still did not get the same

answers as my own calculations, the values differ greatly. they don't even

share the same pattern.

the rdf compute in LAMMPS is in use for a very long time by many people,

and it has been validated against external code many times. i am very

confident, that it is correct. could it be, that your own code is not

implemented correctly, or you are not correctly using the rdf compute in

LAMMPS or that you are not plotting/analyzing the output from LAMMPS

correctly?

> I should also mention that I am doing this on an fcc crystal. does it

make any difference?

no.

axel.