Host galaxy association module by mcoughlin · Pull Request #411 · boom-astro/boom

mcoughlin · 2026-03-12T11:30:52Z

This PR adds a host galaxy association module with DLR + Bayesian scoring.

Based on implementation in https://github.com/alexandergagliano/Prost.

Copilot

Pull request overview

Adds a new host-galaxy association utility (DLR + Bayesian scoring) and wires it into the ZTF/LSST enrichment workers, driven by new config options and LS_DR10 crossmatches.

Changes:

Introduces utils::host with DLR computation, candidate scoring, and association output structs.
Stores host_galaxy association results back into alert documents during ZTF/LSST enrichment when enabled.
Extends config.yaml crossmatch configuration to include LS_DR10 and adds a host_galaxy config block.

Reviewed changes

Copilot reviewed 7 out of 7 changed files in this pull request and generated 6 comments.

Show a summary per file

File	Description
`src/utils/mod.rs`	Exposes new `host` utility module.
`src/utils/host.rs`	Implements host association logic + unit tests.
`src/enrichment/ztf.rs`	Projects `cross_matches`, runs host association, writes `host_galaxy` into Mongo updates.
`src/enrichment/lsst.rs`	Runs host association from `cross_matches`, writes `host_galaxy` into Mongo updates.
`src/enrichment/mod.rs`	Re-exports host association types.
`src/conf.rs`	Adds `host_galaxy` to `AppConfig`.
`config.yaml`	Enables host association and adds LS_DR10 crossmatch configs.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

You can also share your feedback on Copilot code review. Take the survey.

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Copilot · 2026-03-12T11:36:27Z

@mcoughlin I've opened a new pull request, #412, to work on those changes. Once the pull request is ready, I'll request review from you.

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Theodlz

Just wondering, because that is not in the PR description. How is this meant to be used? Right now we crossmatch with NED for example, nothing fancy. I want to figure out how we don't end up with too many places for people to look at potential host galaxies.

First off, since this is purely positional (I suppose?), shouldn't we have that info at the object level? If so, and since the crossmatches need to happen just once per object and we do this in the alert worker, should we move that there? Then in the enrichment worker, we could just compute a hosted boolean feature (ala star, near brightstar) and use that to build the babamul topics for example?

github-actions · 2026-03-12T11:42:58Z

Throughput results (ba81fe4b967d26d96cb4fd8b180c85d5b8d8e432):

New wall time	Baseline wall time	Difference
159.6	183.2	-12.00%

github-actions · 2026-03-12T11:49:29Z

Throughput results (e76f78cc9670e230707c0250031ad3c16c3ba3b8):

New wall time	Baseline wall time	Difference
181.3	183.2	-1.00%

github-actions · 2026-03-12T11:49:55Z

Throughput results (ec18f1edbbcd753c11ceef4840cb8589fcef8038):

New wall time	Baseline wall time	Difference
186.8	183.2	1.00%

github-actions · 2026-03-12T11:50:00Z

Throughput results (11268a9342453f8469060441e32725eb847a0c3a):

New wall time	Baseline wall time	Difference
193.0	183.2	5.00%

github-actions · 2026-03-12T11:50:12Z

Throughput results (806ae0733c076ea3d28a8b80676b928e96763ac9):

New wall time	Baseline wall time	Difference
187.1	183.2	2.00%

github-actions · 2026-03-12T11:50:17Z

Throughput results (b32982720144eeb398a893677da397094962e05e):

New wall time	Baseline wall time	Difference
186.5	183.2	1.00%

#412) Non-finite `shape_e1`/`shape_e2` (NaN, ±∞) would pass through the existing `shape_r` guard and propagate NaN into `axis_ratio`, `b_arcsec`, DLR, and Bayesian posteriors — since `e.min(0.999)` and `f64::max` do not suppress NaN. ## Changes - **Early rejection**: return `None` when `shape_e1` or `shape_e2` are non-finite, before any arithmetic. - **Post-computation guard**: return `None` if derived `q`, `a_arcsec`, or `b_arcsec` are non-finite or ≤ 0 (defensive against edge cases like a NaN `min_b`). - **Tests**: `test_tractor_shape_non_finite_ellipticity` covering NaN `e1`, NaN `e2`, both NaN, `+∞ e1`, `−∞ e2`. ```rust // Before: NaN silently propagated let e = (shape_e1 * shape_e1 + shape_e2 * shape_e2).sqrt(); // NaN if inputs are NaN let e_clamped = e.min(0.999); // NaN.min(0.999) == NaN // After: reject early if !shape_e1.is_finite() || !shape_e2.is_finite() { return None; } // ...and defensively after derivation: if !q.is_finite() || q <= 0.0 || !b.is_finite() || b <= 0.0 { return None; } ```  --- ✨ Let Copilot coding agent [set things up for you](https://github.com/boom-astro/boom/issues/new?title=✨+Set+up+Copilot+instructions&body=Configure%20instructions%20for%20this%20repository%20as%20documented%20in%20%5BBest%20practices%20for%20Copilot%20coding%20agent%20in%20your%20repository%5D%28https://gh.io/copilot-coding-agent-tips%29%2E%0A%0A%3COnboard%20this%20repo%3E&assignees=copilot) — coding agent works faster and does higher quality work when set up for your repo. --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: mcoughlin <2291947+mcoughlin@users.noreply.github.com>

mcoughlin · 2026-03-12T12:03:45Z

@Theodlz The idea is that GHOST gives you a metric to associated transients with hosts that understands morphology. Because most galaxies are ellipses on the sky, shape matters a lot when doing the association.

github-actions · 2026-03-12T12:07:29Z

Throughput results (694708dfaf461f75e6572bb1dc2a3376c2f5f422):

New wall time	Baseline wall time	Difference
187.7	183.2	2.00%

github-actions · 2026-03-12T12:14:08Z

Throughput results (5f0d67a43930498575d778f1fa3a1c07e2c60a17):

New wall time	Baseline wall time	Difference
185.6	183.2	1.00%

github-actions · 2026-03-12T12:17:53Z

Throughput results (471c531c20602b07d28458e5f60e56f0607d8f76):

New wall time	Baseline wall time	Difference
189.2	183.2	3.00%

github-actions · 2026-03-12T12:38:16Z

Throughput results (9f897e58f1f7f7e18b0465a5ce9ff559b85b8a98):

New wall time	Baseline wall time	Difference
196.6	183.2	7.00%

Add host galaxy association module with DLR + Bayesian scoring

126885a

Copilot AI review requested due to automatic review settings March 12, 2026 11:30

Copilot started reviewing on behalf of mcoughlin March 12, 2026 11:31 View session

Copilot AI reviewed Mar 12, 2026

View reviewed changes

Comment thread src/utils/host.rs

Comment thread src/utils/host.rs Outdated

Comment thread src/utils/host.rs Outdated

Comment thread src/utils/host.rs Outdated

Comment thread src/utils/host.rs Outdated

Comment thread src/utils/host.rs Outdated

Update src/utils/host.rs

bff9dad

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Copilot AI mentioned this pull request Mar 12, 2026

Guard tractor_shape_to_ellipse against non-finite ellipticity inputs #412

Merged

mcoughlin and others added 4 commits March 12, 2026 06:36

Update src/utils/host.rs

46b20d7

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Update src/utils/host.rs

d204e50

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Update src/utils/host.rs

27ed951

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Update src/utils/host.rs

3e1e1e3

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Theodlz reviewed Mar 12, 2026

View reviewed changes

Copilot AI and others added 2 commits March 12, 2026 06:54

Move host galaxy association to alert worker

250ba0b

cargo formatting

5c20466

fix tests

85353a3

mcoughlin requested a review from Theodlz March 12, 2026 12:41

Uh oh!

Conversation

mcoughlin commented Mar 12, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI commented Mar 12, 2026

Uh oh!

Theodlz left a comment

Choose a reason for hiding this comment

Uh oh!

github-actions Bot commented Mar 12, 2026

Uh oh!

github-actions Bot commented Mar 12, 2026

Uh oh!

github-actions Bot commented Mar 12, 2026

Uh oh!

github-actions Bot commented Mar 12, 2026

Uh oh!

github-actions Bot commented Mar 12, 2026

Uh oh!

github-actions Bot commented Mar 12, 2026

Uh oh!

mcoughlin commented Mar 12, 2026

Uh oh!

github-actions Bot commented Mar 12, 2026

Uh oh!

github-actions Bot commented Mar 12, 2026

Uh oh!

github-actions Bot commented Mar 12, 2026

Uh oh!

github-actions Bot commented Mar 12, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants