[gmx-users] g_analyze -ee

[Justin A. Lemkul]

dina dusti wrote:
> Dear Gromacs Specialists,
> 
> Sometimes, when I do "g_analyze -f      .xvg -av -ee error.xvg" , I take 
> following warning, and I don't know how to fix it.
> 
> Set   1:  err.est. 0.000596502  a 0.29217  tau1 24.8856  tau2 443.641
> Warning: tau2 is longer than the length of the data (864000)
>          the statistics might be bad
> invalid fit:  e.e. 0.285911  a 0.995985  tau1 1434.16  tau2 1.41932e+09
> Will fix tau2 at the total time: 864000
> Set   2:  err.est. 0.00859257  a 0.995692  tau1 1432.7  tau2 864000
> Set   3:  err.est. 0.00527967  a 0.588024  tau1 804.83  tau2 3603.55
> a fitted parameter is negative
> invalid fit:  e.e. 0.00421461  a 1.08722  tau1 1672.3  tau2 6955.25
> Will fix tau2 at the total time: 864000
> a fitted parameter is negative
> invalid fit:  e.e. -nan  a 1.00449  tau1 1455.53  tau2 864000
> Will use a single exponential fit for set 4
> Set   4:  err.est. 0.00453113  a 1  tau1 1400.34  tau2 0
> 
> Please help me.
> Thanks in advance from your response.
> 

It likely means the data are poorly converged.  Refer to the paper cited in 
g_analyze -h regarding the error calculation for details and a complete 
description of the error estimate method.

-Justin

-- 
========================================

Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
MILES-IGERT Trainee
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================