adding model with interpolated data

[jpdeleon]

Hi, I wonder if it's a good idea to add interpolated model for values with large gaps. For example, TESS data always has large gap mid-way each sector. As shown in the figure below, it would be great if best model prediction during data gap (+ uncertainty) can be superposed after flattening the data e.g. using gp method.

[hippke]

That would be a useful feature. For splines or GPs, it would be easy to implement. For a slider, if the boxcar is shorter than the gap, it would be impossible by definition. In the data you show here, there is a non-zero chance that some methods produce fits that "go to hell" (i.e., +-infinity) in the gap.
As of now, wotan offers no uncertainty estimate. If you need this, I recommend a GP-based tool like celerite.