Bispectra

The Bk folder contains the bispectra measurements. The data is organized following the general hierarchical structure described in Suite folders.

For each simulation of the latin hypercube sets we have computed the bispectrum of the total matter at redshift zero. For the hydrodynamic simulations we have also compute bispectra for the gas and dark matter at redshift zero. We do not compute bispectra of the stellar or black-hole components, due to the small number of particles in those datasets, hence the low signal-to-noise.

Bispectra are computed using two approaches. At low-k, the spectra are computed using an FFT-based approach, as implemented in the Pylians code. This estimates B(k1, k2, mu), where mu is the angle between k1 and k2. Data are assigned to a 128 x 128 x 128 grid, using triangular-shaped-cloud interpolation, before the bispectra are computed for a range of k values up to k = 5 h/Mpc.

At high-k, the FFT-based approach is inefficient, since it requires a large FFT grid. In this case, we compute bispectra using the HIPSTER code, which uses a configuration-space estimator to extend to small scales without computational penalties (strictly via a convolution with a smooth window which is of negligible importance on small scales). This estimates the Legendre multipoles of the bispectrum, B_ell(k1, k2), for a range of ell up to ell = 5 and k from 0 to 50 h/Mpc. Additional details of the method hyperparameters can be found in the bispectrum headers.

The files containing the power spectra are named as:

Pk_type_method_rsd_z=0.00.txt

where type can be m (for total matter), g (for gas), or c (for dark matter). method'' is either ``lowk or highk, indicating the two regimes given above, and rsd specifies the treatment of redshift-space distortions, either unspecified (real-space), or RSDX for distortions added along axis X. The file headers contain a variety of useful information about the sample.

The data is simply given as a set of text files. For the low-k spectra:

import numpy as np

# get the name of the file with the low-k bispectrum (here for CDM in real-space)
f_Bk = 'CAMELS/Bk/IllustrisTNG/LH_456/Bk_c_lowk_z=0.00.txt'

# read the angular bins
mu_arr =  np.loadtxt(f_Bk,max_rows=1)
# read the bispectrum
data = np.loadtxt(f_Bk,skiprows=15)
k1, k2, Bk = data[:,0], data[:,1], data[:,2:]

# k1, k2  contain the wavenumbers in h/Mpc
# Bk contains the bispectrum measurements in (Mpc/h)^6 for each k1, k2 pair (row) and mu bin (column)

Whilst for the high-k spectra:

import numpy as np

# get the name of the file with the low-k bispectrum (here for gas using redshift-space axis 2)
f_Bk = 'CAMELS/Bk/IllustrisTNG/LH_456/Bk_g_lowk_RS2_z=0.00.txt'

# read the bispectrum
data = np.loadtxt(f_Bk)
k1, k2, Bk = data[:,0], data[:,1], data[:,2:]
# load the angular multipoles
ells = np.arange(len(Bk[0]))

# k1, k2  contain the wavenumbers in h/Mpc
# Bk contains the bispectrum multipole measurements in (Mpc/h)^6 for each k1, k2 pair (row) and multipole ell (column).