Skip to content

Data Sources

miniMillennium

  • Box size: 62.5 Mpc/h
  • Snapshots: 64 (z ≈ 127 → 0)
  • Tree format: lhalo_binary (trees_063.0 – trees_063.7)
  • Scale factor file: millennium.a_list
  • SAGE output: model_0.hdf5 with groups Snap_0 – Snap_63

Launch:

visage --par input/millennium.par

microUchuu

  • Box size: 96 Mpc/h
  • Snapshots: 50 (z ≈ 13.9 → 0)
  • Tree format: lhalo_binary (tree_0_0_0.dat)
  • Scale factor file: Uchuu100_scalefactor.txt

Launch:

visage --par input/microuchuu.par

LightSAGE lightcone output

A LightSAGE cli_lightcone run produces a single flat HDF5 file — no periodic box, no per-snapshot groups, one array per field at the root:

  • Pass-through SAGE fields keep their original CamelCase names — Posx/Posy/Posz, StellarMass, SnapNum, Type, Mvir, Rvir, SfrDisk/SfrBulge, …
  • Computed fields are lowercase — ra, dec, distance, redshift_cosmological, redshift_observed, sfr
  • Mass fields are in 10^10 Msun/h (same convention as SAGE HDF5 output); positions are observer-frame comoving Mpc/h
  • Cosmology and provenance are recorded in the SageOutputHeader and LightconeOutputHeader groups

Launch:

visage --lightcone /path/to/lightcone.h5

See Lightcone Mode for what's different about viewing a lightcone versus a SAGE box.

How the par file is used

ViSAGE's parse_par() reads the .par file you pass to the CLI. Relative paths in the par file are resolved relative to the par file's parent directory (your SAGE26 root), so you can run visage from anywhere as long as you give it the absolute or correctly relative path to the par file.

Data that is never committed

Tree files, HDF5 outputs, and scale-factor lists are all listed in .gitignore. ViSAGE only reads them at runtime.