Hi all,
i have a TMultiGraph composed by 50 TGraphErrors with 10 points, each point with a different error on Y axis. A point has the same X value for each TGraph errors. for example for X=0.01 I have 50 points (each from a different tgrapherrors), each one with his error.
I want to perform a fit to the TMultiGraph , does anyone knows how the errors weight are calculated? Because the results I get is quite strange, and I don’t know if having the same X for the points in different graph could be a problem.
Thank you,
Francesco
Hi,
The weight is calculated as the inverse of the square of the error for each point and it should not matter if you have the same X values. If you don’t have errors in X, or you don;t want the error in X to be contributed you can use the option “EX0”.
The error in x (multiplied by the function first derivative) is added in quadrature to the Y error to form a total point error.
Lorenzo