next up previous contents index
Next: License Agreement Up: The fit- background- and Previous: The background functions

Measure functions

     
tv > fit measure definition dy-chi-square  Minimization of χ2 weighted by data errors (Maximization of a gauss distribution).

chi(f)2 = ∑i=0numdata {( {f(Xi) - YidYi} } 2
f:
fit function
X i
channel
Y i
spectrum-value of channel X i
dY i
error of Y i
tv > fit measure definition y-chi-square

Minimization of χ2 weighted by data (Maximization of a gauss distribution).

chi(f)2 = ∑i=0numdata {(f(Xi) - Yi)2|Yi|}
f:
fit function
X i
channel
Y i
spectrum-value of channel X i
tv > fit measure definition poisson   Maximization of a poisson distribution.

P(f) = ∏i=0numdata {f(Xi)YiYi!} ⋅ exp(-f(Xi))
f:
fit function
X i
channel
Y i
spectrum-value of channel X i

The really maximized and printed measure is the function:

C:
i=0numdata -ln(Yi!)

next up previous contents index
Next: License Agreement Up: The fit- background- and Previous: The background functions
Andreas Fitzler
7/13/2000