From 0c1db8de370e2059f933346c5d411cb022f8453b Mon Sep 17 00:00:00 2001
From: sjplimp
Cij(delta) = |Vi(t)*Vj(t+delta)| +Cij(delta) = ave(Vi(t)*Vj(t+delta))which is the correlation value between input values Vi and Vj, separated by time delta. Note that the second value Vj in the pair is -always the one sampled at the later time. The vertical bars represent -an average over every pair of samples in the set that are separated by +always the one sampled at the later time. The ave() represents an +average over every pair of samples in the set that are separated by time delta. The maximum delta used is of size (Nrepeat-1)*Nevery. Thus the correlation between a pair of input values yields Nrepeat correlation datums: diff --git a/doc/fix_ave_correlate.txt b/doc/fix_ave_correlate.txt index 9fe4b5120c..655ef0833e 100644 --- a/doc/fix_ave_correlate.txt +++ b/doc/fix_ave_correlate.txt @@ -106,12 +106,12 @@ output timestep. The initial time could be the beginning of the simulation or the last output time; see the {ave} keyword for options. For the set of samples, the correlation value Cij is calculated as: -Cij(delta) = |Vi(t)*Vj(t+delta)| :pre +Cij(delta) = ave(Vi(t)*Vj(t+delta)) :pre which is the correlation value between input values Vi and Vj, separated by time delta. Note that the second value Vj in the pair is -always the one sampled at the later time. The vertical bars represent -an average over every pair of samples in the set that are separated by +always the one sampled at the later time. The ave() represents an +average over every pair of samples in the set that are separated by time delta. The maximum delta used is of size ({Nrepeat}-1)*{Nevery}. Thus the correlation between a pair of input values yields {Nrepeat} correlation datums: