[gmx-users] Ensemble average in dipole correlation function

[Kim Q Hwan]

Dear all,

I am coding my own correlation analysis program. Coding itself is not problem, but I am confusing how to deal ensemble average properly. My two questions are like this:

0. In ensemble average with molecular dynamics, <A>_ensemble = <A>_time from Ergodic hypothesis, if I have enough frames. Is it right?

1. Dipole correlation is defined as <u(t)*u(0)>. How can I determine u(t)*u(0)? 
 In other words, are u(t) and u(0) averaged separately, like u(t)=average u_i(t) and u(0)=average u_i(0)
                          or  averaged together, like u(t)*u(0) = average u_i(t)*u_i(0), for whole molecule i?

Thank you. 

-- 
Kim Q Hwan
Center of Nano-Liquid
Department of Physics & Astronomy
Seoul National University
Sent with Sparrow (http://www.sparrowmailapp.com/?sig)