FFTPredict :: (SourceModel, GainModel) -> Visibilties

Joshuaalbert / DSA2000-Cal

DSA-2000 Calibration and Forward Modelling

https://www.deepsynoptic.org/overview

MIT License

1 stars 0 forks source link

Closed Joshuaalbert closed 4 months ago

Joshuaalbert commented 6 months ago

Perform visibility prediction using FFT + degridding using IDG.

How:

We incorperate gains into prediction (including w-term) using gridding/degridding.
The prediction process happens via two stages. V_G = FFT(image) -> Degrid(V_G, gain_model)
The gain model needs to be dynamic in shape as each patch from uv plane has different lm coords in the low resolution image space.