Class
NumCosmoGalaxyShapePopGaussLocal
Description [src]
final class NumCosmo.GalaxyShapePopGaussLocal : NumCosmo.GalaxyShapePop
{
/* No available fields */
}
Truncated-Gaussian intrinsic ellipticity distribution, per-galaxy width.
Same truncated-Gaussian family as NcGalaxyShapePopGauss (the “Global”
variant) — a SIBLING, not a subclass: both derive directly from
NcGalaxyShapePop and share the family’s mechanism (data layout,
eval_p, gen) as plain functions reused across the two files (see
nc_galaxy_shape_pop_gauss_private.h), not through inheritance. The only
real difference is where the width $\sigma$ comes from: here it is not a
shared model parameter but is resolved per galaxy from an input RMS
ellipticity e_rms carried on NcGalaxyShapePopData (read from the catalog
column “e_rms”). Since $\sigma$ parameterizes an untruncated Gaussian
while e_rms is the truncated-disk RMS, resolving $\sigma$ from a target
e_rms requires inverting the (monotonic) forward map; this is done once
per galaxy in prepare() via bisection, never in the marginal-likelihood hot loop.
Functions
nc_galaxy_shape_pop_gauss_local_clear
Decreases the reference count of gspgl by one, and sets the pointer
gspgl to NULL.
Instance methods
Methods inherited from NcGalaxyShapePop (10)
nc_galaxy_shape_pop_e_rms
Computes the per-component intrinsic RMS ellipticity
$e_\mathrm{rms} = \sqrt{\langle |\chi_I|^2 \rangle / 2}$ implied by data.
nc_galaxy_shape_pop_eval_p
Evaluates the normalized density $P(x)$ of the squared intrinsic
ellipticity modulus, reading the resolved parameters from data (no live
parameter access).
nc_galaxy_shape_pop_eval_p_rho2
Evaluates the same density as nc_galaxy_shape_pop_eval_p(), but parameterized
directly by $\rho^2$ instead of $x$. The default implementation just
substitutes $x=\rho^2/(1+\rho^2)$ into eval_p(); a subclass overrides it
only when its density has a genuinely more direct or better-conditioned
form in $\rho^2$ (see NcGalaxyShapePopBeta, whose density is a rational
power of $\rho^2$ with no need to ever form $1-x$ by subtraction).
Intended for calculators that already work in $(u,v)$ via a
disc-compactifying substitution (e.g. NcGalaxyShapeFactorQuad), where
$\rho^2$ is the natural coordinate.
nc_galaxy_shape_pop_eval_p_rho2_g_series
Taylor-in-$g$ analog of nc_galaxy_shape_pop_eval_p(): composes this
population’s own fully normalized density with the (already computed,
population-independent) shear-map series $x(g)$, order by order in $g$.
There is no generic default — every subclass used with
NcGalaxyShapeFactorSeriesLensed must provide its own implementation; the
base class errors clearly otherwise.
nc_galaxy_shape_pop_free
Decreases the reference count of gsp by one.
nc_galaxy_shape_pop_gen
Samples an intrinsic ellipticity $\chi_I = e_1 + i e_2$ from data.
nc_galaxy_shape_pop_get_mode_x
Gets the mode of $x=|\chi_I|^2$ for this galaxy’s resolved population
density, i.e. where $P(x)$ peaks. This is a per-data capability
(data->ldata_get_mode_x), not a virtual method, mirroring
nc_galaxy_shape_pop_get_sigma(); unlike that capability, 0 is always a
meaningful default (the model is assumed radially symmetric about
$\chi_I=0$ unless it says otherwise), so this never errors.
nc_galaxy_shape_pop_get_sigma
Gets the (untruncated) Gaussian width sigma resolved for this galaxy, for
models that parameterize their density this way (the Gauss family, Global
or per-galaxy). This is a per-data capability (data->ldata_get_sigma),
not a virtual method: any concrete model sharing the same internal
representation can populate it, regardless of its position in the class hierarchy.
nc_galaxy_shape_pop_prepare
Resolves the model parameters (and the per-galaxy data->e_rms, for per-galaxy
models) into data: the subclass-specific parameters stored in data->ldata
that nc_galaxy_shape_pop_eval_p() / nc_galaxy_shape_pop_eval_p_rho2() read.
nc_galaxy_shape_pop_ref
Increases the reference count of gsp by one.
Properties
Properties inherited from NcmModel (9)
NumCosmoMath.Model:implementation
NumCosmoMath.Model:name
NumCosmoMath.Model:nick
NumCosmoMath.Model:params-types
NumCosmoMath.Model:reparam
NumCosmoMath.Model:scalar-params-len
NumCosmoMath.Model:sparam-array
NumCosmoMath.Model:submodel-array
NumCosmoMath.Model:vector-params-len
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.