absolute flux and M2

[artus at obiwan.kri.physik.uni-muenchen.de]

Hello Kristian,

 > This looks fine, but I think I understand your confusion. A common
 > mistake (at least I made it many times while working on McStas) is to
 > think of the weights as probabilities, ie. numbers smaller than
 > one. Weights are NOT probabilities and they may well be greater than
 > one. For example, for the Source_flux example above, N=384 and I=127625,
 > so the average neutron weight is a bit more than 300.
 > 
 > In fact, we can see that between the two sets of numbers, I increses by
 > a factor of about 10**8, while M2 increses with a factor of about
 > 10**16 = (10**8)**2. This fits well since M2 is the sum of the squares
 > of the weights (I = sum(p_i) and M2 = sum(p_i**2)). You still get a good
 > estimate of the statistical error from the square root of M2.
 > 
 > Does that clarify matters for you?

Yes, thanks. My error was to ignore that sigma**2 is calculated. I somehow
supressed that ...

 > When you say 'transform using solid angle', I am not quite sure what you
 > mean; the Source_flux component itself makes a correction for the solid
 > angle in which neutrons are emitted ... ?
 > 

The solid angle in my example 

COMPONENT source = Source_flux(
        radius = 0.050,
        dist = 2.000,
        xw = 0.040, yh = 0.075,
        E0 = 81.804,    
        dE = 1.636,     
        flux = 3.63E+12)

is 2.387E-4. 0.04 * 0.075 are exactly width an height of my guide. If
according to my example I(flux)=3.66E+6 and I(flat)=0.0496 then

I(flux) = I(flat) * flux * solid angle / 4 / pi

(The numbers don't fit exactly because I didn't use the same seed for
the two calculations.)

Georg