Validation run plan

This page documents the planned validation runs for the four stable estimators (SMF, wp(rp), w(θ), ΔΣ(rp)) and for the in-development LF estimator. Each section lists the dataset, the launch script, the expected output, and the quantitative pass criterion.


Uchuu HOD mocks — SMF, WPRP, WTHETA

Purpose: verify that the estimators recover the truth on a controlled N-body + HOD realisation where the input two-point statistics are known.

Item

Details

Dataset

Uchuu N-body halo catalogue with an HOD applied to match BGS clustering

Estimators

SMF (1/Vmax), wp(rp), w(θ)

Launch scripts

scripts/measure_joint_sumstat.py --survey mock --stats SMF WP WTHETA

Output

One HDF5 file per stellar-mass threshold; figures from scripts/plot_joint_sumstat.py

Pass criterion — SMF

Integrated number density within 20% of the HOD input at each mass bin

Pass criterion — WPRP

Amplitude and slope consistent with the HOD projected correlation within jackknife errors on scales 0.1–20 Mpc

Pass criterion — WTHETA

Same, on scales 1–100 arcmin

Status

Scripts exist; runs not yet formally documented as a validation pass


Uchuu HOD mocks — DeltaSigma

Purpose: verify the ESD pipeline end-to-end on a mock where the true projected matter density profile is known from the HOD.

Item

Details

Dataset

Uchuu HOD galaxy catalogue (lenses) + synthetic source catalogue with shear computed from the HOD NFW profiles

Estimators

ΔΣ(rp) with jackknife covariance

Launch scripts

scripts/measure_joint_sumstat.py --survey mock --stats ESD_HSC (substitute mock survey flag when available)

Output

ΔΣ(rp) vs rp; comparison to NFW model expectation

Pass criterion

Recovered ΔΣ within jackknife errors of the analytic NFW expectation on scales 0.1–10 Mpc; no systematic offset at > 2σ

Status

Not yet run; synthetic source catalogue not yet generated


COSMOS — SMF and LF

Purpose: validate the SMF (stable) and LF (development) estimators against the widely-used COSMOS2015 measurements of Ilbert et al. (2013).

Item

Details

Dataset

COSMOS2015 galaxy catalogue (Laigle et al. 2016), photo-z + stellar masses

Estimators

SMF 1/Vmax, SWML; LF 1/Vmax per band (development)

Launch scripts

scripts/measure_joint_sumstat.py --survey cosmos --stats SMF LF

Reference

Ilbert et al. (2013) (SMF); Loveday et al. (2015) (LF)

Output

φ(M★) per redshift bin; overlay with Ilbert+ 2013 data points

Pass criterion — SMF

Δφ/φ < 30% in bins with > 20 galaxies (accounting for cosmic variance); Schechter α within 0.2 of Ilbert+ 2013 values at z < 1

Pass criterion — LF

(development) Δφ/φ < 30% vs Loveday+ 2015 in r-band bins

Status

Scripts exist; figure script at docs/scripts/make_literature_figures.py; formal pass/fail not yet documented


GAMA — SMF and WPRP

Purpose: validate SMF and projected clustering against GAMA, which provides a well-characterised low-redshift galaxy sample.

Item

Details

Dataset

GAMA DR3 galaxy catalogue (Driver et al. 2022); spectroscopic z, stellar masses

Estimators

SMF 1/Vmax; wp(rp) with jackknife covariance

Launch scripts

scripts/measure_joint_sumstat.py --survey gama --stats SMF LF WP

Reference

Baldry et al. (2012) (SMF); Farrow et al. (2015) (WPRP)

Output

φ(M★) vs Baldry+ 2012; wp(rp) vs Farrow+ 2015 per mass bin

Pass criterion — SMF

Δφ/φ < 20% in bins with > 50 galaxies; Schechter M★ within 0.15 dex of Baldry+ 2012

Pass criterion — WPRP

wp within jackknife errors of Farrow+ 2015 on scales 0.1–20 Mpc

Status

SMF script exists; WPRP script needs --survey gama support; not yet run


BGS (DESI Legacy Survey) — full joint suite

Purpose: the primary science target; validate all four stable estimators jointly on the BGS spectroscopic sample.

Item

Details

Dataset

DESI BGS Bright (EDR or Y1); stellar masses from SED fitting; HSC Y3 / DES Y3 / KiDS sources for lensing

Estimators

SMF, wp(rp), w(θ), ΔΣ(HSC + DES + KiDS); joint jackknife covariance (100 regions)

Launch scripts

scripts/measure_joint_sumstat.py --survey bgs --stats SMF WP WTHETA ESD_HSC ESD_DES ESD_KIDS

Reference

Wright et al. (2017) (SMF proxy); GAMA WPRP as low-z reference

Output

Joint HDF5 file per mass threshold; figures from scripts/plot_joint_sumstat.py

Pass criterion

All four estimators consistent with overlapping GAMA/COSMOS measurements at z < 0.3; internal consistency between the three ESD surveys within jackknife errors

Status

Production runs ongoing (see logs/measure_mstar*.log); formal validation comparison not yet automated