Maybe not? But take BLAS: on my Mac, uv's numpy uses newaccelerate, which is significantly better than openblas, and also the best you can get from conda. But on this random Unix machine I'm connected to, uv's numpy uses scipy-openblas, and I can do better with a specialised implementation.
There are also packages where the conda version is significantly faster, due to better compiled dependencies. But uv is definitely a step in the right direction generally!
A new 2025.2 release of GLASS, working towards Array API compatibility (incl. jax).
This release moves the GLASS to the Cosmology API. That means you can easily run with a background cosmology provided by CAMB, Astropy, or your own classes through a common API standard.
github.com/glass-dev/gl...
New Publication at the Open Journal of Astrophysics: "Shot noise in clustering power spectra" by Nicolas Tessore (University College London, UK) and Alex Hall (University of Edinburgh, UK)
doi.org/10.33232/001...
Exactly right! The puzzle is to invent something that makes sense
This has been puzzling me. Given a set of redshift distributions nᵢ(z), i = 1, 2, ..., it's easy to work out the variance σᵢ² for each bin. How to fill in the rest of the "covariance matrix" so that it has intuitive properties? E.g. perfectly overlapping bins having σᵢⱼ = σᵢ² = σⱼ².
Play and score! (Bit unfair perhaps)
Figure 1
Figure 2
Figure 3
Shot noise in clustering power spectra. Nicolas Tessore et. al. https://arxiv.org/abs/2507.03749
Separating the point made from the DESI results, I still think it’s good practice to test any such expansion against a higher-order one (like DESI effectively did with the GP). That could be as simple as a likelihood-ratio test or so.
Great results from DESI, but not nearly enough commentary about what phantom dark energy means in terms of GR
yeah, it’s an old-fashioned drain unblocker
aren't you supposed to add water, sugar, a few dashes of bitter, and an orange slice?
Fantastic little book for World Book Day from DK and Lego
First 2025 release of GLASS is out: github.com/glass-dev/gl...
It comes with a big change that allows you to easily create FLASK-like simulations from a set of n(z) without simulating the entire lightcone. This isn’t usually the best way to go, but can sometimes be.
We’d probably just waste the money on postdocs’ livelihood instead
I feel like an idiot! Someone just showed there is a "web" tab for Google results that displays the plain old Google results from yesteryear!
BlueSky etc. are fine, but I wish we also had a dedicated space for professional discourse in Cosmology, Astrophysics, Astronomy, etc.
Something like categorytheory.zulipchat.com, which seems phenomenally useful to me.
Is the Vatican City the country with the highest number of astronomy papers/capita?
That’s a shame. I’m not following social media all that much but I enjoy, for example, reading mathematical discussions over at mathstodon.xyz from time to time. I was hoping there’d be the equivalent for our field somewhere.
Fair enough, I was hoping for science content from scientists!
Are there feeds with science content, instead of feeds with random content by scientists?
Hey @ojastro.bsky.social, do you accept "research notes" as understood by RNAAS? (short text, single figure or table, possibly no sections, etc.)
We need to go back to this plot style
Much better, thanks!
Sad to see the state of Bluesky after a long hiatus.
Far too much galaxies, not nearly enough cosmology.
PSA: Numpy is currently deprecating and planning to remove the trapz() function in v2.0, due to appear in a month or so. This sounds like utter madness to me, since it's probably one of the most common operations I use.
As it turns out, this mysterious undocumented 𝑀-inversion parameter is really a 𝚖𝚒𝚗𝚟𝚎𝚛𝚜𝚒𝚘𝚗 parameter.
I find it very interesting that the publications describing the famous Hershey fonts are fairly easy to find, but the original data files seem lost to time. I bet that it will be the same 40 years hence, and papers about code survive much longer than the codes themselves.
I finally had enough trying to homogenise BibTeX files from multiple authors, and started on a no-dependencies, pure Python formatter for BibTeX:
pip install tidybib # requires python>=3.10
tidybib main.bib # reformats main.bib
Next step: detect duplicate entries with different keys.
