Add toy galaxy dataset for training and testing

CosmoStat / autometacal

Metacalibration and shape measurement by automatic differentiation

MIT License

4 stars 1 forks source link

Add toy galaxy dataset for training and testing #3

Closed EiffL closed 3 years ago

EiffL commented 3 years ago

To run the various tests we will be interested in, we will need some ready available galaxy images, under a series of noise and PSF conditions. This issue is to track the development of a TensorFlow Dataset class that uses GalSim to generate a dataset of postage stamps, with each record having the following entries:

unsheared/unconvolved galaxy image
noise statistics
PSF

This way, we will be able to add a random shear to the galaxy image on the fly, and add whatever amount of noise we want.

EiffL commented 3 years ago

@b-remy can you link here the example of the kappaTNG dataset you implemented?

andrevitorelli commented 3 years ago

https://github.com/CosmoStat/jax-lensing/blob/master/jax_lensing/datasets/kappatng.py

EiffL commented 3 years ago

Ok, so actually, looks like this issue is becoming redundant with #4 , I'm going to close it and modify the other one