Interpolate SN lightcurves with GPs


We want to interpolate SN lightcurves with gaussian processes. The interpolation will be used to construct templates, which are then used to construct bolometric lightcurves, filling in for missing data.

The lightcurves have diverse sampling, diverse noise, generally a smooth behavior, with more variability at early times, rather than late times (after the \({}^{64}\mathrm{Ni}\) decay starts dominating).

Here is an example of a very good, very well sampled lightcurve, with photometry from 2 telescopes, which is not consistent within the tiny tiny, probably underestimated observational uncertainties:

1993J V band, two sources
Here is an example of a not well sampled lightcurve, with relatively largeerrorbars