incorporate input covariates in transit model

[jpdeleon]

Is it possible to incorporate additional input covariates e.g. centroid offset, airmass, etc for joint systematics+transit modeling? I have ground-based light curves that have strong in-transit trend that can be explained by telescope pointing offset. Although using e.g. Matern 32 baseline model will fit the trend, it seems reasonable to use physically-motivated covariates to properly take into account the systematics. Is this added complexity necessary at all?

[MNGuenther]

Glad that you ask @jpdeleon! We developed allesfitter in the spirit of Agile Software Development, i.e. it is modular with room for these expansions if users request it. Since noone has asked about it yet, it is not yet in there ;) It will be some time until we get around to this, but can certainly include it in the future. For now, I would recommend you to fit the GP on the systematics time series (centroid, airmass, or the like) beforehand, and then use the GP's time scale (not amplitude) as a prior for the photometry fit.