/direct_ia_theory

CosmoSIS modules for IA modelling

Primary LanguagePython

Direct IA Theory

Code for IA modelling. The modules here are designed to be plugged into a CosmoSIS pipeline.

Contact: Simon Samuroff (s.samuroff@northeastern.edu)

Contents:

util

add_1h_ia:

  • Summary: Reads pre-computed one and two halo power spectra from the block, and add them together.
  • Language: python
  • Inputs: 1h and 2h IA power spectra P^1h_GI(k,z), P^2h_GI(k,z), P^1h_II(k,z), P^2h_II(k,z)
  • Outputs: Combined 1h+2h IA power spectrum P_GI(k,z), P_II(k,z)

flatten_pk:

  • Summary: Reads a pre-computed power spectrum and integrates along the redshift direction. Also applies galaxy bias and IA amplitudes if desired.
  • Language: python
  • Inputs: IA power spectra P_GI(k,z),P_II(k,z); redshift distributions p_1(z),p_2(z)
  • Outputs: Flattened spectra P_GI(k),P_II(k)

promote_ia_term:

  • Summary: Reads a power spectrum, renames it.
  • Language: python
  • Inputs: Generic power spectrum P(k,z)
  • Outputs: The same power spectrum, but now called something else in the data block P(k,z)

likelihood

add_1h_ia:

  • Summary: Computes a likelihood for some combination of wgg, wg+ and w++ 2pt data. Scale cuts and ordering specified in the params file.
  • Language: python
  • Inputs: Theory correlations wgg, wg+, w++; data vector wgg', wg+', w++'; covariance matrix C.
  • Outputs: A likelihood and a chi2.

power_spectra

schneider_bridle:

  • Summary: Computes IA power spectra using the fitting fuctions of https://arxiv.org/abs/0903.3870.
  • Language: python
  • Inputs: None
  • Outputs: One halo intrinsic alignment power spectra P^1h_GI(k,z), P^1h_II(k,z)

projection

projected_corrs_limber:

  • Summary: Reads a flattened power spectrum and performs a Hankel transform with the appropriate Bessel function to generate wg+, wgg, w++ etc.
  • Language: C
  • Inputs: Flattened power spectra P(k)
  • Outputs: Real space projected correlations as a function of perpendicular separation wg+(r_p), w++(r_p), wgg(r_p).

projected_corrs_legendre:

  • Summary: Reads an IA power spectrum computes the line-of-sight projected correlations using Legendre polynomials.
  • Language: python
  • Inputs: IA power spectra P_GI(k,z), P_II(k,z).
  • Outputs: Real space projected correlations as a function of perpendicular separation wg+(r_p), w++(r_p), wgg(r_p).

photometric_ias:

  • Summary: Computes projected IA correlations in the case where one or both of the samples have finite redshift uncertainty (i.e. for photometric samples) using the prescription of https://arxiv.org/abs/1008.3491. Note that this process involves (a) computing C_ells via a series of Limber integrals, (b) a Hankel transform, and then (c) projecting in Pi and redshift. All of these steps are done internally (might take a few seconds, depending on settings), making use of CCL for (b).
  • Language: python
  • Inputs: IA power spectra P_GI(k,z), P_II(k,z).
  • Outputs: real space projected correlations as a function of perpendicular separation wg+(r_p), w++(r_p), wgg(r_p).