Fitting dLorentz/dx


jhtilgner
01-11-2010, 01:11 PM
I have to solve fitting the derivative of the lorentz distribution.

In funcsG the derivative of lorentz distribution is notated as

H := a[k+0];
P := a[k+1];
B := a[k+2];

PmI := P-I;

Hh2 := sqr( H );
Ph2 := sqr( P );
Bh2 := sqr( B );

PmIh2 := sqr( PmI );

MyV := Bh2 + PmIh2;

y := y + 2 * Bh2 * H * PmI / sqr( MyV );

dyda[k+0] := 2 * Bh2 * PmI / sqr( MyV );

dyda[k+1] := 2 * Bh2 * H * ( Bh2 - 3 * PmIh2 ) / ( sqr(MyV) * MyV );

dyda[k+2] := 4 * B * H * ( ( PmIh2 - Bh2 ) * PmI ) / ( sqr(MyV) * MyV );

But there is no fitting. The default values for peakintensity, position and width are returned. The values of input data are
positive and negative too.

Why?