AHF halo catalogues, protohalo shape data, and merger tree cross-match files for CDM and WDM N-body cosmological simulations (L=25 Mpc/h, N=512^3)Data products for Mostoghiu Paun et al. (2025)Data products for "Tidal adaptive softening and artificial fragmentation in cosmological simulations"B25_N512_CDM_3LPT, B25_N512_WDM_3LPT, WDM_CDM_crossmatch, CDM_WDM_crossmatch

Full description

This dataset contains the simulation data products generated for the study of tidal adaptive force softening and artificial fragmentation in cosmological N-body simulations, published as:   Mostoghiu Paun R. A., Croton D., Power C., Knebe A., Ussing A. J., Duffy A. R.   (2025). Tidal adaptive softening and artificial fragmentation in cosmological   simulations. MNRAS 542, 735–746. https://doi.org/10.1093/mnras/staf1229 --- Simulations --- Two cosmologies (CDM, WDM) × two initial redshifts (z_ini = 39, 99) × two softening schemes (fixed, tidal adaptive) = 8 simulations, each labelled z{39,99}_{fixed,adapt}. The data are also used as the basis for a machine learning spurious halo classifier, available at: [here: spurious_halo_classifier] --- Contents --- Four zip archives are provided: B25_N512_CDM_3LPT.zip AHF halo catalogues (*.AHF_halos) and protohalo shape files (SPHERICITY_ALL_INFO.*.hdf5) for cold dark matter (CDM) simulations across four configurations: z39_fixed, z39_adapt, z99_fixed, z99_adapt, corresponding to initial redshifts z_ini = 39 or 99 and fixed or tidal adaptive force softening. B25_N512_WDM_3LPT.zip Same structure as CDM, but for 1 keV warm dark matter (WDM) simulations. WDM_CDM_crossmatch.zip AHF MergerTree cross-correlation files (.croco format) with WDM haloes as primaries and CDM haloes as secondaries. Used as the source of spurious halo labels in the ML classification project. CDM_WDM_crossmatch.zip AHF MergerTree cross-correlation files with CDM haloes as primaries and WDM haloes as secondaries (i.e. the reverse direction). These files enable a mutual-best-match (bijective) consistency check on the primary label. Note: the crossmatch files were produced specifically for the ML classification project and were not part of the analysis in Mostoghiu Paun et al. (2025). The paper's spurious halo identification relied on protohalo sphericity, not CDM–WDM counterpart matching. --- Usage --- Unpack all four archives under data/raw/ following the layout in config.yaml; the classifier's DuckDB database is then built from scratch (make gold). See the repository README for the full pipeline. --- Simulation setup --- Gravity code: GIZMO (Hopkins 2015, 2017) Box size: L = 25 Mpc/h Resolution: 512^3 particles; m_p = 1e7 M_sun/h Cosmology: Planck18 (H_0=67.74, Omega_m=0.31, Flat LCDM) Initial conditions: 3LPT with MONOFONIC (Michaux et al. 2021) Initial redshifts:  z_ini = 39 (late-start, preferred) and z_ini = 99 WDM model: 1 keV (Viel et al. 2005 transfer function) Softening schemes:  fixed (eps = 1.22 kpc/h) and tidal adaptive (Hopkins et al. 2023, min eps = 122 pc/h) Halo finder: AHF (Gill et al. 2004; Knollmann & Knebe 2009) Merger trees: MERGERTREE (packaged with AHF)

Notes

External Organisations
Swinburne University of Technology; Universidad Autónoma de Madrid

Associated Persons
Robert Adriel Mostoghiu Paun (Creator); Darren Croton (Contributor); Alexander Knebe (Contributor); Adam Ussing (Contributor); Alan Duffy (Contributor)