Skip to main content

PSF_FIT_METHOD

Constant PSF_FIT_METHOD 

Source
pub const PSF_FIT_METHOD: &str = "weighted least squares on the bin energies (not their logarithm), \
     model = exact annulus integral of E f(r), weights 1/sigma_i^2 from per-history errors, bins with \
     sigma_i = 0 left out; start from a log grid over the widths with linear least squares for the \
     amplitudes; Levenberg-Marquardt on the logarithms of the parameters with the analytic Jacobian; \
     covariance (J^T W J)^-1 over the fitted parameters in their natural units, not scaled by the \
     reduced chi^2; E fixed to the slab energy (default) or fitted, see normalization";
Expand description

How the fit is done, as recorded in a PsfFit.