Connor Stone, PhD

@entaylor oh I really like the idea of specifying the way that the parameters are handled (log, normal, arcsinh, etc) in the parameter specification! I'll have to think very carefully...

Hi @entaylor , indeed there is a lot of bookkeeping to do! It's so cool to see built out pipelines with the code :) I think there are a bunch...

Hi @entaylor , I am working on some example scripts to help users get started. One of the scripts takes an image and a segmentation map from the user and...

Thanks @entaylor ! For AstroPhot, the primary fitting algorithm is Levenberg-Marquardt which, when things are going well, does actually treat the model as linear and therefore should land on the...

Ah, this is interesting. Do you think it is possible to ignore the flux parameters entirely until the end? Or would they need to be updated at each iteration? I...

Ah, I understand now. That would be really easy to implement for a single model even now. But for group models (and even more for joint models) that would be...

I've been thinking about this some more. I now think that in the LM algorithm this is all handled automatically. The Hessian encodes these linear relationships, so they are already...

This seems like more of a hack to speed up convergence. Something that would be relevant along these lines would be the ability to break up an image into sub-windows...

@dr-guangtou good question, there is a `warp` model class which can have a variable shape with radius and has essentially as much expressiveness as an isophotal ellipse model. However the...

Playing around with the warp model I have found it is actually able to represent spiral arm and bar features very well, possibly better than elliptical isophotes. So it is...