Releases: ismael-mendoza/bliss
Releases · ismael-mendoza/bliss
Zenodo Release for OJA Publication (v1)
What's Changed
- active tests and update packages by @ismael-mendoza in #1
- Ensure consistent performance on dim blends (no galaxies in padding) by @ismael-mendoza in #3
- Detection encoder performance by @ismael-mendoza in #4
- train autoencoder by @ismael-mendoza in #5
- add stars with cutoff by @ismael-mendoza in #7
- prepare to train deblender by @ismael-mendoza in #8
- Train deblender2 by @ismael-mendoza in #9
- remove concat output transform by @ismael-mendoza in #11
- Incorporate best pl practices by @ismael-mendoza in #12
- Debug deblender by @ismael-mendoza in #14
- Update packages 07 20 24 by @ismael-mendoza in #15
- New ellipticity measure by @ismael-mendoza in #16
- use boostrap to compute errors on mean/median of residuals by @ismael-mendoza in #17
- Ensure AE is trained to convergence by @ismael-mendoza in #18
- changes the loss function of the deblender to be a sum not a mean by @ismael-mendoza in #37
- removes the misguided idea of creating an snr catalog by @ismael-mendoza in #41
- add test related to checking angles are in degrees by @ismael-mendoza in #42
- add more testing on tiled catalog by @ismael-mendoza in #43
- reorganize notebooks into folders by @ismael-mendoza in #44
- add code to create final blending dataset (w/ correct galaxy density) by @ismael-mendoza in #45
- checks individual galaxy dataset is ready to roll by @ismael-mendoza in #48
- implement system to track runs and datasets by @ismael-mendoza in #49
- I/23 by @ismael-mendoza in #50
- fixes the loss for encoders by @ismael-mendoza in #51
- notebooks added that demonstrate overfitted deblender estimating flux jointly correctly by @ismael-mendoza in #53
- move io to lsst for catlaog/mags by @ismael-mendoza in #55
- add early stopping to 3 encoders by @ismael-mendoza in #58
- adds trained deblender that performs reasonably well by @ismael-mendoza in #59
- add avg. stellar density by @ismael-mendoza in #65
- retrain all encoders with correct star density by @ismael-mendoza in #66
- implement toy figure tweaks by @ismael-mendoza in #70
- use background subtracted images by @ismael-mendoza in #75
- Update python poetry by @ismael-mendoza in #77
- make detection results by @ismael-mendoza in #78
- Organize notebooks and update poetry 2.0 by @ismael-mendoza in #82
- new results in toy figure by @ismael-mendoza in #84
- additional deblender figures by @ismael-mendoza in #88
- sampling fig draft by @ismael-mendoza in #89
- use uv by @ismael-mendoza in #92
- check odd size ptiles are possible by @ismael-mendoza in #95
- Encoders ingest ptiles for training by @ismael-mendoza in #96
- new sampling figure and cleanup by @ismael-mendoza in #98
- Rerun all results and add reproducibility scripts by @ismael-mendoza in #99
- modifications to figures from round2 review by @ismael-mendoza in #103
- final toy figures by @ismael-mendoza in #104
- ensure color blind friendly figures by @ismael-mendoza in #105
New Contributors
- @ismael-mendoza made their first contribution in #1
Full Changelog: https://github.com/ismael-mendoza/bliss/commits/desc-oja-v1