# [gmx-users] correlation function

Jamie Seyed jamie.seyed at gmail.com
Sun Aug 23 16:30:51 CEST 2009

```Hi Justin,

On Sat, Aug 22, 2009 at 9:17 PM, Justin A. Lemkul <jalemkul at vt.edu> wrote:

>>
>> On Sat, Aug 22, 2009 at 8:44 PM, Justin A. Lemkul <jalemkul at vt.edu<mailto:
>> jalemkul at vt.edu>> wrote:
>>
>>
>>
>>    Jamie Seyed wrote:
>>
>>        Hi Justin and Mark,
>>        Thanks. Actually I made a mistake and it seems going to the zero
>>        (even negative!)
>>        --But I wanted to know how I can write my command to get only
>>        one graph (average of all instead of all of them). I read the
>>        manual chapter-8 and by average I mean I should get one curve
>>        starting at 1 and coming down to zero...I also tried
>>
>>
>>    That's not an average; that's the shape of the ACF curve.
>>
>>  Yes I need to get that shape using g_analyze..!.
>>
>
> Right, I just don't understand why you think you need to obtain some sort
> of average.  What do you mean by "all of them"?  How many graphs are there?

Actually when I open autocorr.xvg I can see from data/feature extraction/
the last set is S1295 (N=101, autocorr.xvg) which means they are 1296
graphs each with poor statistics start at 1 and coming down. To have a ACF
for a system I expect to get one graph with good statistic... I hope I make
it clear now.

>
>
> <snip>
>
> No, only ACF based on coordinates. I read from the list that people said I
>> can use g_rotacf  -d for calculation of correlation functions...and actually
>> I did it for a simple system. But for new system using another index file it
>> is complaining.
>>
>>
> Then I don't think g_rotacf is the tool you want.  There is no magic way to
> simply calculate some sort of general correlation function.  You calculate
> an ACF from the specific data you have.  So if you're interested in some
> sort of position-dependent data, then rotational properties may or may not
> be relevant.
>
> You can certainly use g_rotacf if you think it will be meaningful; see
> below regarding your .tpr file.
>
> <snip>
>
> tpbconv  -s f.tpr  -f f.trr  -n f.ndx  -o tpxout.tpr
>>
>>
>
> Don't use -f.  If you do, tpbconv will think you are trying to create a new
> .tpr file that starts off from the time frame present in the .trr file.

Yes I removed -f and I got  tpxout.tpr and used it for g_rotacf
g_rotacf  -f f.xtc  -s tpxout.tpr  -n f.ndx  -o rotacf.xvg  -d
but the result I got is like y=2x that starts at (0,0) which y=C(t) and
x(Time,ps). Which is very strange... I don't understand what are these
results using this command. (I should mention that f.xtc contains the
coordinates of selected atoms). Many thanks for your attention and your
help/Jamie

>
>
> -Justin
>
>
> --
> ========================================
>
> Justin A. Lemkul
> Ph.D. Candidate
> ICTAS Doctoral Scholar
> Department of Biochemistry
> Virginia Tech
> Blacksburg, VA
> jalemkul[at]vt.edu | (540) 231-9080
> http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin
>
> ========================================
> _______________________________________________
> gmx-users mailing list    gmx-users at gromacs.org
> http://lists.gromacs.org/mailman/listinfo/gmx-users
> Please search the archive at http://www.gromacs.org/search before posting!
> Please don't post (un)subscribe requests to the list. Use the www interface
> or send it to gmx-users-request at gromacs.org.