Method
NumCosmoGalaxySDObsRedshiftGausseval_pz_given_zp
Declaration [src]
gdouble
nc_galaxy_sd_obs_redshift_gauss_eval_pz_given_zp (
NcGalaxySDObsRedshiftGauss* gsdorgauss,
gdouble z
)
Description [src]
Evaluates the conditional true redshift distribution for a photometric bin.
This distribution gives the probability that a galaxy observed in the photometric
redshift bin $[z_{p,\mathrm{min}}, z_{p,\mathrm{max}}]$ has true redshift z:
$$
P(z|z_{p,\mathrm{min}}, z_{p,\mathrm{max}}) = \frac{P(z) \, W(z_{p,\mathrm{min}},
z_{p,\mathrm{max}}|z)}{N}
$$
where $W$ is the Gaussian integral over the photometric redshift bin:
$$
W = \int_{z_{p,\mathrm{min}}}^{z_{p,\mathrm{max}}} \mathrm{Gauss}(z_p|z,\sigma_z(z))
\, \mathrm{d}z_p,
$$
with $\sigma_z(z) = \sigma_0 (1 + z)$, and $N$ is the normalization constant.
The distribution is computed once and cached, so subsequent calls with the same parameters will be very fast. The cache is invalidated when the model state changes.
This is particularly useful for computing the redshift distribution of galaxies in weak lensing tomographic bins defined by photometric redshift cuts.