<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
On 12/08/2011 12:30 AM, Jinan Niu wrote:
<blockquote
cite="mid:1a416d0e.cfa3.131b94053e6.Coremail.njn0516@163.com"
type="cite">
<div style="line-height: 1.7; color: rgb(0, 0, 0); font-size:
14px; font-family: arial;">
<div>On 8/08/2011 11:08 PM, Mark Abraham wrote:<br>
<br>
>I don't quite follow your question, but for correlation
functions to be <br>
>computed accurately, you need data collected much more
frequently >than <br>
>the relevant time scales. That can require good planning.<br>
>Mark<br>
</div>
<div>Hi,Mark and other gmx users：</div>
<div>when I used g_velacc to calculate VAF, I found the range
of frequency of the power spectrum is related with dt, so if
is there is some solutions? can the complication of gmx by
double decision solve it?<br>
</div>
</div>
</blockquote>
<br>
If the value for g_velacc -dt is comparable to the correlation time
scale of the system, then the results of an analysis will be
sensitive to that value. In such cases, you need to use the smallest
possible -dt value (i.e. omit the value) and/or collect data more
frequently.<br>
<br>
Mark<br>
</body>
</html>