Reporting fit statistics

[phajy]

I wonder if there is an easy way to calculate model-dependent fit statistics such as reduced $\chi^2$ values, for example. $\chi^2_\nu = \chi^2 / \text{d.o.f.}$ where the $\chi^2$ value is derived from the data and the fit (model folded through response), and the number of degree of freedom (d.of.) is equal to $(\text{number of data bins}) - (\text{number of free parameters})$ which come from the data and model, respectively.

[fjebaker]

Yes we can add a function that does that. The number of free parameters is already tracked, and the number of data bins is simply count(mask) on the data mask, so it's just about adding that to the report card of the fit! Good idea!