Class
NumCosmoHaloBiasCastro
Description [src]
final class NumCosmo.HaloBiasCastro : NumCosmo.HaloBias
{
/* No available fields */
}Linear halo bias calibrated together with the Castro halo mass function.
The peak-background split prediction obtained from NcMultiplicityFuncCastro,
$$ b_\mathrm{PBS} = 1 - \frac{1}{\delta_c}\frac{\mathrm{d}\ln f}{\mathrm{d}\ln\nu}, $$
multiplied by a calibrated correction that depends on $\Omega_m(z)$, on the slope
$s = \mathrm{d}\ln\sigma_R/\mathrm{d}\ln R$ and on the clustering amplitude $S_8$,
$$ b = b_\mathrm{PBS}\,A_0\,(1 + a_1\Omega_m(z))\,(1 + b_1 s + b_2 s^2)\,(1 + c_1 S_8). $$
The derivative is the total one along the mass direction, which is what the correction was calibrated against. It is not obtained by differentiating an interpolation: the model is closed-form in $\sigma_R$ and $s$, so the two partial derivatives are taken by perturbing those numbers directly through nc_multiplicity_func_castro_eval_full(), and the chain rule closes with $\mathrm{d}s/\mathrm{d}\ln R$ obtained from the transform itself.
The mass function this bias is attached to must use a NcMultiplicityFuncCastro,
since the correction coefficients were fitted alongside that multiplicity function.
See the Castro Halo Mass Function and Bias theory page, and Castro et al. (2024).
Constructors
nc_halo_bias_castro_new
Creates a new NcHaloBiasCastro with the calibrated correction coefficients.
The mass function mfp must use a NcMultiplicityFuncCastro.
Functions
nc_halo_bias_castro_clear
Atomically decrements the reference count of biasf by one. If the reference count
drops to 0, all memory allocated by biasf is released. Set the pointer to NULL.
Instance methods
nc_halo_bias_castro_S8
Clustering amplitude $S_8 = \sigma_8\sqrt{\Omega_{m,0}/0.3}$, with $\sigma_8$ the present-day variance on $8\,h^{-1}\,\mathrm{Mpc}$. NumCosmo works in $\mathrm{Mpc}$, so the radius is converted with the reduced Hubble parameter.
nc_halo_bias_castro_correction
Calibrated correction multiplying the peak-background split prediction, $A_0(1 + a_1\Omega_m(z))(1 + b_1 s + b_2 s^2)(1 + c_1 S_8)$.
nc_halo_bias_castro_free
Atomically decrements the reference count of biasf by one. If the reference count
drops to 0, all memory allocated by biasf is released.
nc_halo_bias_castro_pbs
Peak-background split prediction
$b_\mathrm{PBS} = 1 - \delta_c^{-1}\,\mathrm{d}\ln f/\mathrm{d}\ln\nu$, with the
derivative taken at fixed slope. This is the partial derivative: for a scale-free
spectrum it is already the total one, and nc_halo_bias_eval() adds the slope-running
term where the spectrum is curved.
Methods inherited from NcHaloBias (4)
nc_halo_bias_eval
Computes the Halo Bias at a given redshift. The mass lnM identifies the scale
at which sigma was evaluated; non-universal models use it to query further
properties of the filtered power spectrum through the mass function. Universal
models ignore it.
nc_halo_bias_free
Atomically decrements the reference count of bias by one. If the reference count drops to 0,
all memory allocated by bias is released.
nc_halo_bias_integrand
This function is the integrand of the mean bias, i.e., the product of the mass function with the bias function. As both functions depend on the standard deviation of the matter density contrast, we implement this function to compute \f$ \sigma (M, z) \f$ just once.
nc_halo_bias_peek_mass_function
Gets the mass function this bias was built on.
Properties
NumCosmo.HaloBiasCastro:A0
Overall amplitude of the correction to the peak-background split prediction.
Properties inherited from NcHaloBias (1)
Signals
Signals inherited from GObject (1)
GObject::notify
The notify signal is emitted on an object when one of its properties has its value set through g_object_set_property(), g_object_set(), et al.