[gmx-users] using g_analyze for calculating distance autocorrelation functions

[申昊]

Dear GROMACS users,

I have been working on calculating distance autocorrelation functions by using g_analyze.
The codes i used was 
g_analyze -f dist.xvg -ac autocorr.xvg -temp 300.
'dist.xvg' is about the average distance against simulation time. The two columns are both positive datas.
However, the result show a mixture of positive and negative datas. In my opinion, the autocorrelation function should be always positive. 

Can anyone help me with that ?

HaoShen