[gmx-developers] g_energy output compared to g_analyze

[David van der Spoel]

Hi,

the development version of g_energy outputs an error estimate:
Energy                      Average   Err.Est.       RMSD  Tot-Drift
-------------------------------------------------------------------------------
#Surf*SurfTen                267.87        9.9     1551.2   -65.2551 
(kJ/mol)

Running g_analyze on the otuput xvg file gives:
SS1   2.757495e+02   1.555144e+03   1.639265e+01      -0.027    0.005


Back Off! I just backed up /tmp/errest.xvg to /tmp/#errest.xvg.1#
a fitted parameter is negative
invalid fit:  e.e. 24.2959  a 1.13429  tau1 0.336301  tau2 1.20464


Is there a relation between the error estimate in g_energy and in 
g_analyze? I would expect that the same algorithm would be used, or does 
g_energy use the intermediate results that are stored in the edr file in 
a clever way that g_analyze can not?

-- 
David van der Spoel, Ph.D., Professor of Biology
Dept. of Cell & Molec. Biol., Uppsala University.
Box 596, 75124 Uppsala, Sweden. Phone:	+46184714205. Fax: +4618511755.
spoel at xray.bmc.uu.se	spoel at gromacs.org   http://folding.bmc.uu.se