This code encapsulates two methods to estimate the r_c, r_t
parameters of a King (1962) profile.
The first one is described in Pieres et al. (2016), and corresponds to maximizing the King profile likelihood through brute force applied on a grid. This method is divided into two: 'KP_2' and 'KP_4'. The former only fits the core and tidal radius, while the latter also fits the eccentricity and rotation angle.
The second one I developed for ASteCA and estimates an "optimal"
radius based on maximizing the (normalized) difference between the
number of members within the cluster region, and the number of field
stars. When testing it here, I discovered that this method returns
a fraction of the real r_t value. This fraction is ~0.7 (maybe 1/sqrt(2)?)
for all values of CI if we average the K_cp factor (concentration
parameter), and varies from ~0.22 to ~0.42 for the K_cp, averaging over
the CI.
- Estimate the field density and number of members parameters from the data at each step, instead of fixing them.
- Make the center coordinates also free parameters?