Relating the structure of dark matter halos to their assembly and environment

Calculators for Dark Matter Halo Structural Properties

We provide both an on-line calculator and a Python programming interface for users to predict halo structural properties (i.e., concentration, shape, and spin) from other halo properties (i.e., halo mass, assembly history, tidal field and bias factor). The detail description of the dimension reduction and regression methods can be found at the Appendix A of the paper. The output of the calculator is the structural parameter synthesized from a mean value (output from the regressor of template data) combined with a random noise.

The online calculator only provides limited set of predictors, and can only compute ONE halo structural property for ONE halo at each query. To compute on a large set of halos, or to use more complete and flexible combinations of predictors, please try the Python interface.

$ pip install haloprops
                

from haloprops.structure import StructurePredictor

# fit template data
sp = StructurePredictor()
sp.fit('concentration')

# make prediction for 3 halos from their four properties ( the first PC of assembly history, halo mass, tidal anisotropy parameter, halo bias )
X = [[ 1.55278125e+00,  2.78270723e+04,  1.88321865e-01, 4.02548960e+00],
[ 3.08029618e-01,  2.88465363e+02,  1.81991570e-01, -1.14664588e-01],
[ 3.46576614e-01,  1.57965881e+03,  4.11933281e-01, 2.77537420e+00]]
concentration = sp.predict(X).val
print(concentration)

[8.62213406 8.98157622 7.72968138]

Catalogs of Halo Properties

Here we provide two catalogs extracted from ELUCID simulation (L=500, N=3072). The first is 'Highmass Sample', which includes all of the halos in the simulation in the top-hat mass range $[10^{13.4}, 10^{14.6}) h^{-1}M_\odot$. The second is 'Lowmass Sample', which includes a subset of halos in the simulation in the top-hat mass range $[10^{11}, 10^{13.4}) h^{-1}M_\odot$. The selection of the second sample is that, we first divide the mass range $[10^{11}, 10^{13.4}) h^{-1}M_\odot$ into bins whose width are $0.3\ {\rm dex}$, and then randomly choose 10000 halos in each bin. The union the two samples are the sample $\rm S_L$ defined in our paper.