coding-galaxy-gen-references

name	coding-galaxy-gen-references
description	External reference resources for the galaxy sim - real-galaxy catalogs, N-body library design references, Astropy primitives worth mirroring. Use when the sim needs a real-world parameter or validation. Triggers - NASA ADS, SIMBAD, NED, astropy, FITS, sky survey, REBOUND, GADGET, n-body simulation, galaxy catalog, SDSS, GAIA, JWST, Hubble.

References for galaxy-gen

Repo: galaxy-gen. Sister skills: coding-galaxy-gen-astrophysics (galaxy-scale physics), coding-galaxy-gen-cosmology (universe-scale physics).

This skill is about pointers, not theory. Use it when the sim needs a number from the real universe, or when you want to look at how a battle-tested gravitational simulator handles a problem instead of redesigning it yourself.

When to reach for real data

Three honest cases for a procedural sim:

• Picking defaults. "What is a reasonable disk scale length for an Sb spiral?" 3.5 kpc. Could just hardcode it, but if you want a range, real catalogs have one.
• Sanity-checking output. "Does my generated rotation curve look like NGC 3198's?" Plot real curves over your sim curves. Plenty published.
• Validation gallery. "What does a real Sbc galaxy look like?" SDSS gallery, Hubble Legacy Archive, JWST. Free, indexed.

You almost never want to ingest real FITS files into a procedural sim. The astrophysics-data-guide content is for analysis pipelines, not generators.

Literature lookup: NASA ADS

The bibliographic database for astronomy. Indispensable for "what does the literature say about X galaxy property". Free API, 3000 requests/day.

• Register for a token - https://ui.adsabs.harvard.edu/user/settings/token (free, no expiry).
• Auth - Authorization: Bearer $ADS_API_TOKEN header. Keep the token out of the repo (env var, secret store, or .env listed in .gitignore).
• Search endpoint - GET https://api.adsabs.harvard.edu/v1/search/query?q=...&fl=title,author,year,bibcode,citation_count&rows=10&sort=citation_count+desc.
• Object search - q=object:"M31" returns papers about a specific galaxy. Useful when copying real-galaxy parameters into the sim.
• Docs - https://ui.adsabs.harvard.edu/help/api/

For galaxy-gen specifically, ADS is overkill for routine work. It is the right reach when picking sim defaults from published surveys (e.g. "what is the Schechter function's M_star value in the local universe").

Catalogs and image archives

When picking real galaxies to study as visual reference, or copying parameters from observed populations.

• SDSS (Sloan Digital Sky Survey) - https://skyserver.sdss.org/ - the biggest public galaxy catalog. Photometry, spectra, morphology. Galaxy Zoo classifications are derived from SDSS.
• GAIA - stellar catalog for the Milky Way. 1.8 billion stars with parallax, proper motion, color. Use for Milky Way-shaped scenes specifically.
• NED (NASA Extragalactic Database) - https://ned.ipac.caltech.edu/ - the canonical reference for any named galaxy. Multi-wavelength photometry, redshift, classification.
• SIMBAD - http://simbad.u-strasbg.fr/simbad/ - same idea but covers all object types. Better for stellar / Galactic objects.
• Hubble Legacy Archive - https://hla.stsci.edu/ - reference imagery. The pretty pictures.
• JWST archives - https://archive.stsci.edu/missions-and-data/jwst - higher-resolution successors to Hubble. Especially relevant for high-z morphology if the sim renders early-universe galaxies.

Astropy: what is worth knowing

The Python astronomy library. Mostly irrelevant for a Rust/WASM sim, but a few primitives are worth recognizing when comparing notes.

• astropy.coordinates - the standard for sky-coordinate transformations (ICRS, Galactic, ecliptic, alt-az). If the sim ever maps a real galaxy onto its sky, Astropy is the reference implementation.
• astropy.units - dimensional analysis with units. The right model for any "physical-units" type in the Rust side (kpc, M_sun, km/s, etc.). Don't port it; just mirror the convention.
• astropy.cosmology - canned LCDM cosmology. cosmology.lookback_time(z), cosmology.luminosity_distance(z). Same integrals the sim needs in coding-galaxy-gen-cosmology. Cross-check sim lookup tables against Astropy output once.
• FITS - the standard data format. Skip unless ingesting real catalogs. The sim does not need to emit FITS.

N-body simulation libraries: design references

If galaxy-gen ever moves from "static sampling" to "time-evolved gravitational dynamics", the wheel has been built many times. Read the design, do not necessarily port it.

• REBOUND - https://rebound.readthedocs.io/ - C library with Python bindings. Symplectic integrators (WHFast, IAS15) optimized for celestial mechanics. Small-N, high-accuracy. The right reference if galaxy-gen ever simulates stellar orbits in a galactic potential.
• GADGET-4 - https://wwwmpa.mpa-garching.mpg.de/gadget4/ - the production cosmological N-body + SPH simulator. Tree + particle-mesh for gravity, SPH for gas. Massive scale (millions to billions of particles). The reference for "how would the pros do this".
• AREPO - moving-mesh hydro alternative to GADGET. Used for the IllustrisTNG simulation suite. Visually striking but probably overkill for a browser sim.
• Galpy - https://github.com/jobovy/galpy - pure-Python galactic-dynamics library. Lighter than GADGET, focused on stellar orbits in analytic potentials. Closest in scope to what a browser galaxy sim would do.

Useful patterns to lift from these:

• Tree codes (Barnes-Hut) - O(N log N) gravity. Threshold the box-opening angle theta; cells too far to resolve get treated as point masses. Maps cleanly to a WASM implementation.
• Particle-mesh (PM) - FFT-based. O(N log N) via Poisson solve on a grid. Right for cosmological-scale sims, overkill for single-galaxy.
• Symplectic integrators - preserve energy over long integrations. Leapfrog is the simplest; KDK (kick-drift-kick) is the standard. Use over RK4 for anything orbiting.

Computational-physics framing for galaxy-gen

Beyond N-body specifically, the broader comp-physics methods inventory has a few relevant entries:

• Particle-in-cell (PIC) - if galaxy-gen ever models galactic-scale magnetic-field-driven plasma (probably not for a procedural sim).
• SPH (smoothed-particle hydrodynamics) - gas dynamics as particles. Used in GADGET. Relevant if the sim renders gas/dust as physical, not as a texture.
• Adaptive mesh refinement (AMR) - grid-based hydro that refines near density gradients. Different paradigm; not needed for a particle-only sim.

For a procedural sim that fakes most of this with sampling + visual approximations, treat these as a reference vocabulary for any future "make it more realistic" pivot.

When this skill stops being useful

The moment galaxy-gen commits to a real-time interactive procedural sim with a fixed visual budget (browser, single galaxy in view), most of the above becomes overkill. The astrophysics + cosmology skills are the active references. Reach here only when picking defaults, validating output, or designing a dynamics layer.

Sources

• wentorai/physics-skills - the bundle the adjacent content came from. Includes astrophysics-data-guide (Astropy/FITS), nasa-ads-api (full API guide), and computational-physics-guide (MD/MC/N-body taxonomy).
• NASA ADS, SDSS, GAIA, NED, SIMBAD, Hubble Legacy, JWST archive - linked inline above.
• REBOUND, GADGET-4, AREPO, Galpy - linked inline above. None pulled into the repo; these are reference targets.