Weighted linear fit

Hi,

I have a linear distribution with different errors for x and y.
So I want to make a weighted linear fit with respect to the error weight.
At the moment I use “pol1” but I think it is not the right one(?)
What do I have to use?

Cheers,
Gordon

Hi,
you can create a TGraphErrors from your x,y and errorX and errorY values and then call TGraphErrors::Fit(“pol1”)

Lorenzo

Hi,

thanks for answer!

Cheers,
Gordon