Skip to content

feat: jit(fit_from) round-trip for Delaunay source + MGE lens #461

@Jammy2211

Description

@Jammy2211

Overview

Extend Path A (jax.jit-wrapped analysis.fit_from) to FitImaging with a Delaunay pixelization source AND an MGE lens bulge, so that jax.jit(analysis.fit_from)(instance) returns a fully-populated FitImaging matching the NumPy-path scalar. This is the intersection of fit_imaging_pytree_delaunay (#453) and the MGE PoC.

Plan

  • Scaffold delaunay_mge_pytree.py combining MGE bulge + Isothermal + ExternalShear lens with Hilbert image-mesh + Delaunay pixelization source (single pixelization, exercises the narrow fallback from PyAutoGalaxy#361).
  • Run under use_jax=True; iterate any unregistered types.
  • Expected: no new registrations if predecessors landed clean.
Detailed implementation plan

Affected Repositories

  • autolens_workspace_test (primary)
  • PyAutoLens, PyAutoGalaxy, PyAutoArray — only if unregistered types surface

Branch Survey

Repository Current Branch Dirty?
PyAutoLens main clean
PyAutoGalaxy main clean
PyAutoArray main clean
autolens_workspace_test main clean

Suggested branch: feature/fit-imaging-pytree-delaunay-mge
Worktree root: ~/Code/PyAutoLabs-wt/fit-imaging-pytree-delaunay-mge/

Key Files

  • autolens_workspace_test/scripts/jax_likelihood_functions/imaging/delaunay_mge_pytree.py — new

Original Prompt

See admin_jammy/prompt/issued/fit_imaging_pytree_delaunay_mge.md.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions