Conversation
…e change
Regenerate dataset/point_source/simple/ seed data and update the
expected_likelihood literals in scripts/jax_likelihood_functions/point_source/
{image_plane,point,source_plane}.py to match the realistic position-noise
scale (0.005") introduced in 931a381. The committed seed dataset was last
regenerated under the old 0.2" noise, so fresh simulator runs were
producing likelihoods that did not match the hardcoded constants.
- image_plane.py / point.py: 1.313508 → -83.38049778
- source_plane.py vmap: -199.1555813 → -331481.25978149
- source_plane.py eager: -199.1555813 → -331481.26508536364
The 1664x magnitude jump on source_plane.py is consistent with chi-squared
rescaling under a 40x position-noise reduction (1/sigma² scaling).
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Rebaseline three JAX point-source likelihood literals after the realistic noise change in 931a381 (
positions_noise_map0.2" → 0.005"). The committed seed dataset was last regenerated under the old noise scale, so any fresh simulator run was producing likelihoods that did not match the hardcoded constants. This commit regenerates the seed dataset and updates the literals so both canonical-checkout runs and clean re-simulations pass.Scripts Changed
scripts/jax_likelihood_functions/point_source/image_plane.py— rebaselineexpected_likelihoodfrom1.313508to-83.38049778scripts/jax_likelihood_functions/point_source/point.py— rebaselineexpected_likelihoodfrom1.313508to-83.38049778(identical toimage_plane.py; both scripts load the same dataset and evaluate at the same prior medians)scripts/jax_likelihood_functions/point_source/source_plane.py— rebaselineEXPECTED_VMAP_LOG_LIKELIHOOD_SOURCE_PLANEto-331481.25978149andEXPECTED_EAGER_LOG_LIKELIHOOD_SOURCE_PLANEto-331481.26508536364. The 1664x magnitude jump matches chi-squared rescaling under a 40x position-noise reduction (1/sigma² scaling).Also regenerates
dataset/point_source/simple/{point_dataset_positions_only,tracer}.jsonso the on-disk seed dataset, the simulator code, and the literals are consistent. Without the regeneration, anyone who deletes the local dataset (forcingshould_simulateto fire) would hit the same failure on next run.Test Plan
autolens_workspace_test/tmp/cluster_c_*_final.log)🤖 Generated with Claude Code