fix: standardize translate fill parameter

[tylerbessire]

Summary

• normalise translate parameters and accept legacy fill_value
• unify heuristics and complete DSL on fill naming and cast color mapping to ints
• add regression test for translate parameter aliasing

Testing

• pip install scipy -q
• pytest tests/test_translate_fix.py tests/test_recolor_fix.py tests/test_beam_search.py -q
• pytest tests -q
• python tools/train_guidance_on_arc.py --epochs 1

---

https://chatgpt.com/codex/tasks/task_e_68c489916a30832286cc8ff5ea78d65c

Summary by CodeRabbit

• Bug Fixes

  • Resolved parameter mismatches in recolor and translate to prevent training failures and warnings.
  • Standardized translate parameters (dy, dx, fill) with backward-compatible support for legacy fill_value.
  • Improved color handling consistency in detection heuristics.

• Tests

  • Added unit tests ensuring translate’s legacy alias works and detection returns canonical parameters.

• Documentation

  • Updated AGENTS.md with Step 4.3 updates detailing fixes for recolor and translate, including test confirmations and compatibility notes.

[coderabbitai]

Caution

Review failed

The pull request is closed.

Walkthrough

Adds parameter canonicalization and alias support across translate and recolor operations, updates caching to use canonical params, adjusts translate signatures and operation registry, refines heuristics output to canonical names, and adds tests and docs entries confirming backward-compatible handling of legacy aliases.

Changes

Cohort / File(s)	Summary
Docs updates AGENTS.md	Appends two Step 4.3 update logs documenting fixes for recolor mapping and translate fill alias; notes test runs.
DSL core canonicalization & translate aliasing arc_solver/dsl.py	Adds _canonical_params to normalize legacy keys (color_map→mapping, fill_value→fill); updates apply_op to canonicalize before cache-keying; extends op_translate(dy, dx, fill, *, fill_value=None) and routes resolved fill to translate_grid.
DSL complete translate API change arc_solver/dsl_complete.py	Changes translate signature/order to (dy, dx, fill, *, fill_value=None); sets fill from fill_value if provided; updates implementation to use fill; updates get_operation_signatures() to ['dy','dx','fill'].
Heuristics canonical output & int coercion arc_solver/heuristics_complete.py	Emits translate ops with canonical params {'dy': dy, 'dx': dx, 'fill': 0}; coerces colors to int in mapping inference.
Tests for translate alias & detection tests/test_translate_fix.py	Verifies legacy fill_value accepted in program application and heuristics detect canonical fill with dy/dx.

Sequence Diagram(s)

sequenceDiagram
  autonumber
  participant Caller
  participant DSL as dsl.apply_op
  participant Canon as _canonical_params
  participant Ops as op_translate
  participant Grid as translate_grid

  Caller->>DSL: apply_op("translate", params={dy, dx, fill_value})
  DSL->>Canon: canonicalize("translate", params)
  Canon-->>DSL: params={dy, dx, fill}
  DSL->>Ops: op_translate(a, dy, dx, fill)
  Ops->>Grid: translate_grid(a, dy, dx, fill)
  Grid-->>Ops: result
  Ops-->>DSL: result
  DSL-->>Caller: result

  note over Canon,Ops: Legacy aliases normalized before caching and dispatch.

Loading

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~25 minutes

Possibly related PRs

• fix: normalize recolor mapping for training #12 — Also introduces parameter canonicalization and cache-key normalization for operation aliases (e.g., recolor mapping), overlapping in approach and affected layers.

Poem

I hop through params, tidy and small,
Turning fill_value to fill for all.
Dy, then dx, we slide with grace—
A grid waltz, neatly in place.
Cache now knows one canonical song,
Recolor, translate—nothing goes wrong. 🐇✨

✨ Finishing touches

• 📝 Generate Docstrings

🧪 Generate unit tests

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch codex/outline-arc-training-and-evaluation-process-ngi3f3

---

📜 Recent review details

Configuration used: CodeRabbit UI

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between cb3a2b9 and c18f264.

📒 Files selected for processing (5)

• AGENTS.md (1 hunks)
• arc_solver/dsl.py (4 hunks)
• arc_solver/dsl_complete.py (2 hunks)
• arc_solver/heuristics_complete.py (2 hunks)
• tests/test_translate_fix.py (1 hunks)

Tip

👮 Agentic pre-merge checks are now available in preview!

Pro plan users can now enable pre-merge checks in their settings to enforce checklists before merging PRs.

• Built-in checks – Quickly apply ready-made checks to enforce title conventions, require pull request descriptions that follow templates, validate linked issues for compliance, and more.
• Custom agentic checks – Define your own rules using CodeRabbit’s advanced agentic capabilities to enforce organization-specific policies and workflows. For example, you can instruct CodeRabbit’s agent to verify that API documentation is updated whenever API schema files are modified in a PR. Note: Upto 5 custom checks are currently allowed during the preview period. Pricing for this feature will be announced in a few weeks.

Please see the documentation for more information.

Example:

reviews:
  pre_merge_checks:
    custom_checks:
      - name: "Undocumented Breaking Changes"
        mode: "warning"
        instructions: |
          Pass/fail criteria: All breaking changes to public APIs, CLI flags, environment variables, configuration keys, database schemas, or HTTP/GraphQL endpoints must be documented in the "Breaking Change" section of the PR description and in CHANGELOG.md. Exclude purely internal or private changes (e.g., code not exported from package entry points or explicitly marked as internal).

Please share your feedback with us on this Discord post.

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}