cheymans / kids1000_3x2pt

Paper draft for 3x2pt letter

kids1000_3x2pt

This paper was published in Heymans & Troester et al 2021: https://arxiv.org/abs/2007.15632 This repository keeps all the plotting scripts that were used in the paper. Please feel free to use and modify them for your analyses.

kids1000_3x2pt's Issues

Add A_bary to monster param plot

As marginalising over baryons is something that we do, and DES doesn't, we should probably add this parameter to the monster cosmo-param plot.

To make space for it, we could remove the sigma_8 line, as sigma_8 is going to be featured in the main money-plot discussed in #2

3x2pt plot and table labelling

Current version of the plots/tables have
Cosmic shear: EE
Clustering: w
GGL: nE
2x2pt Combinations: EE+w etc
3x2pt Combinations: EE+nE+w

In the methodology paper we have
Cosmic shear: CS
Clustering: GC
GGL: GGL
2x2pt Combinations: CS+GGL etc
3x2pt Combinations: 3x2pt

We could also use the equation notation
Cosmic Shear: C_{\epsilon \epsilon}
Clustering: \xi_{gg}
GGL: C_{n \epsilon}

For consistency with Joachimi et al though, I propose we stick with their convention (CS, GC, GGL)?

Can we quote an upper limit on h?

Our lower limit on h hits the prior with the marginal probability at the lower prior edge exceeding 20% of the peak probability. This is therefore unconstrained.

I wonder could we quote an upper limit for this parameter though? Looks to be around 0.73, which is potentially interesting as this is the Riess number...

Where to include the Planck contours...

Combined constraints are currently plotted in two formats: sigma_8/S8/omega_m, and then the monster plot with everything.

Can I propose monster plot stays the same. (possibly also include another one with the systematics tests as that looked great).

For the "money plot", how about instead showing sigma_8/omega_m/h?
BOSS gives us some constraining power on h, and people are even more interested in h-tensions than they are in S8 tensions.
You'd then zoom into this one - cutting off the edges of the cosmic shear only banana, and then limit the overlay to

• cosmic shear only
• BOSS only
• 3x2pt
• Planck

Update GGL plot

From TT: I noticed that you cut the L1-S2 bin for the GGL data plot. We only cut bins L1-S1, L2-S1, L2-S2, and L2-S3.

IA constraints for CS+GC

Morning @tilmantroester !

I wanted to say that 3x2pt had tighter constraints on A_IA than CS+GC. When we unblind and re-run the CS+GC chain, we shouldn't forget to re-add it to Table C1.

DES/KV450 Summary Plot

@tilmantroester - how easy would it be to make a sigma_8/omega_m plot with

1. KV450+BOSS
2. DES-Yr1 - Abbott et al
3. Planck
4. Our result

This would be for the conclusion/summary section, and allows us to visualise changes from KV450->K1000, and K1000 relative to DES-Yr1 (and in your follow-up paper we can have an updated Figure with DES-Yr 3).

As far as I know there aren't any HSC 3x2pt results, so we wouldn't need to add that.