[gmx-users] g_dipoles, average not the same as from g_analyze

[Emanuel Peter]

Dear gromacs users,

I have following question:

With the use of the g_dipoles tool I have analysed the average dipole moment using following command:

g_dipoles -f *.trr -s *.tpr -n *.ndx -av average.xvg 

I think the option -av calculates the average of the dipole moment of the system I have chosen.

Following output is given:

x1 = time (ps)
y1 = 
y2 = ^2
y3 =  - ^2
...

I have taken the root of y2 (y2^(1/2)) to get  and plotted that against x1.

Mtot.xvg corresponds to the following quantities:

x1 = time (ps)

y1 = M_x
y2 = M_y
y3 = M_z
y4 = M_tot

I plotted M_tot against time and calculated the average by using g_analyze.
Now I have following problem:
My average value from g_analyze is not equal to the value of  which I received from g_dipoles.
The difference between the two values is at about 20 Debye. 
Could you tell me the reason why there exists a difference?
Does g_dipoles calculate the average in another way than g_analyze?

Thank you for your kind answer in advance.

Best regards,

Emanuel Peter