正在加载视频...
视频加载失败
This would have helped me SO MUCH if I had something like this when starting out in astronomy. Took me a few hours to think about how I'd wish this was designed, and then another ~5 hrs for Opus 5.5 to implement it.
28,436 次观看 • 10 天前 •via X (Twitter)
11 条评论

View from a phone screen! Here we have the galaxy SFR - stellar mass sequence, with galaxy image cutouts, as well as colored lines showing the relation for galaxies based on subselections in Dn4000 break strength.

Oh yeah, I forgot to demo my favorite mode: a grand tour over the galaxy manifold!

By popular demand -- you can filter out objects colored by missing values!

You'll probably love the todo list of additional axes: - more common morphological parameters (CAS, G-M20) - environmental graph's top eigenvalues - sequencer x optical spectrum - UMAP-ed image embeddings

Omggg love it

Would love to hear your thoughts on making it better!

congratz for the amazing work sir :)

beautiful! any physics in those horizontal band gaps, or data/presentation artifact?

Nope sorry those are just model fit artifacts in metallicity! I may add a small bit of scatter to smooth then out like Tremonti et al 2004 did.

Oh wow

One of the reasons I've become a believer of the hypothesis that we're undergoing a genuine revolution is that coding is no longer a boundary for scientists like you. Thanks for posting this.
