interpreting the "valid" flag of a fit

[govoni]

Hello iMinuit experts,

I am running a simple fit excercise that seems to me pretty good and still is flagged with my-minuit.valid == false.
Where do I find a reference describing the reasons for a fit success, and how to play with iMinuit configurations so to have a handle on when a fit is marked as invalid?

The fitting snippet is this and the fitting function is a fourth order polynomial:

least_squares = LeastSquares (x_points, y_points, y_err, phi)
my_minuit = Minuit (least_squares, t0 = 1., t1 = 1., t2 = 1., t3 = 1., t4 = 1.) # starting values for m and q
print (my_minuit.valid)
my_minuit.migrad () # finds minimum of least_squares function
my_minuit.hesse () # accurately computes uncertainties

Cheers,

pietro