Getting started¶
Installation¶
pip install . # from the repository root
# or, for development:
pip install -e .
Minimal required packages (installed automatically): fitsio, numpy,
scipy, matplotlib, picca.
Optional:
pyevtk— to export a tomographic box to VTK for 3D visualisation.- SaclayMocks — used by
lelantos.boxdm; a vendored fallback copy ships underlelantos/saclaymocks/, but installing SaclayMocks directly is recommended.
The tomographic solver¶
To run the reconstruction you must place the executable of the tomographic
algorithm in lelantos/lelantos/exec/. The only algorithm currently supported
is Dachshund (a Wiener-filter
reconstruction), expected as dachshund.exe.
Two ways to run¶
1. The all-in-one interface¶
Copy and edit scripts/interface_example.ini (a fully commented reference is
provided as scripts/interface_explanatory.ini), then run:
lelantos_interface.py interface_example.ini
Each stage is switched on/off by a boolean in the [main] section. See the
Pipeline & config page for the meaning of every option.
2. Individual launchers¶
The launchers/ folder holds standalone scripts that expose one stage each.
Edit the variables at the top of the file and run it directly, e.g.:
python launchers/lelantos_convert_delta.py
python launchers/lelantos_void_finder.py
The scripts/ folder additionally provides small argparse / config-parser
utilities to inspect or modify tomographic objects (print a property file, cut
a QSO/DLA catalog, show void statistics, …).
Coordinate & unit conventions¶
- RA / Dec are handled in radians internally (degrees at the plotting boundary), comoving distances and box sizes in Mpc·h⁻¹, velocities in km/s and wavenumbers in h/Mpc.
- The sky ↔ cartesian transform is selected per run by
coordinate_transform(middle,full_angle/full, orcartesian). Most downstream code assumes themiddletangent-plane transform. - The fiducial cosmology enters only at the delta-conversion step, through
Omega_m.