report: name the real gap (destination binding for Report); record why GroupAddr::Pair is not the pivot model by AdaWorldAPI · Pull Request #1331 · AdaWorldAPI/lance-graph

AdaWorldAPI · 2026-10-05T08:14:14Z

Documentation only. This PR first lowered every two-ordinal report to Quack's GroupAddr::Pair (one pass, hi · stride + lo). That was the wrong model and the change has been reverted. What remains is a corrected doc comment and the board record. Runtime behaviour is unchanged.

Why the Pair path was wrong

GroupAddr::Pair addresses the dense product of two dimension domains. Lowering to it automatically made |A| × |B| the execution geometry only because the primitive existed; no consumer had asked for a dense cube. For IAM-sized dimensions (64k × 64k) that is 4·10⁹ slots.

The intended model is PowerShell-Hashtable / Excel-Pivot:

a row's aggregate destination is resolved;
the row folds directly into it;
accumulator state scales with the demanded / resolved destination universe.

GroupAddr::Pair remains valid only for a consumer that explicitly demands a dense mixed-radix destination universe.

The missing capability: destination binding for Report

The missing capability is destination binding / coordinate resolution for Report. Once a compact destination ordinal or functional reference is bound, the existing Local / Via keyed fold is already the execution primitive.

semantic coordinate(s)
        ↓
BIND / RESOLVE DESTINATION
        ↓
compact destination ordinal / functional route
        ↓
existing GroupAddr::Local / GroupAddr::Via keyed fold  (caller-sized K sink, K = destination count)

What CoordSpec cannot express today. Its only variants are Field, Bucket and MaskSet. It has no way to represent:

a functional-reference coordinate, such as user → department;
a composed functional route;
several semantic coordinates bound to ONE compact destination ordinal.

So Report partitions instead.

Law: a join or pivot used only to determine an aggregate destination must compile to destination binding / resolution + fold. It must not compile to an intermediate relation, and must not automatically become a dense product.

How destinations get bound:

Normal path: destinations are bound beforehand. That means a resident ordinal, a CAM / codebook binding, or a Via or composed functional reference.
Fallback only: a query-local compact interner, used solely for a destination universe that genuinely cannot be bound beforehand. It is not the canonical path.
Not in this PR: no new type, address, CoordSpec variant or substrate primitive is introduced.

What the source shows (recon, VERIFIED-IN-CODE)

Keyed reduction. ndarray, mask-risc and Quack share one keyed-reduction walker. It has three addresses: Lane, Via and Pair. All three write a caller-sized K-slot sink.

The sparse path's cost. Report's Layout::Sparse is reached in two stages:

One discovery population scan for each ordinal partition dimension.
T × FoldStates population folds, where T is the product of the partition radices:
- for an ordinal dimension, its discovered members;
- for a Bucket / MaskSet dimension, its full configured radix.

On top of that comes one selection-materialization scan when the ReusedMask carrier is chosen. For example, 100 × 100 observed members × 4 FoldStates = 40,000 fold scans after discovery. This partition-rescan architecture is what destination binding exists to eliminate.

This PR

exec.rs module doc (doc lines only) names the gap as destination binding, as described above.

Board entry entries/2026-10-05-report-pair-key-and-stack-recon.md, titled "Report Pair experiment superseded; destination-binding gap and stack recon":

The Pair cut is kept as a ⊘ superseded record. The greedy-Pair optimization is a historical observation only, not open work.
The sparse cost is corrected as above.
Decision 1 is struck. Decisions 2–5 of the stack recon stand.

Verification

Report unchanged. The planner files (plan.rs, explain.rs, result.rs, lib.rs), Cargo.toml and the tests are byte-identical to main. exec.rs has zero non-doc changed lines.
Checks: lance-graph-report tests pass; fmt and clippy -D warnings are clean.
Board gates: the append-only and entries-index gates pass, and SUPERSESSION was regenerated. citation_decay reports the same output as before this change.

The fold contract and the keyed partial-sink merge live in #1332, not here.

🤖 Generated with Claude Code

https://claude.ai/code/session_01DCEP2fZdHYdMCtcEVcTpS2

coderabbitai · 2026-10-05T08:14:22Z

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check. Docstring coverage is scoped to functions touched by this diff. Analyzed 0 functions across 1…
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately names the destination-binding gap for Report and explains why GroupAddr::Pair is not the pivot model. It is specific and related to the main changes.

Autopilot · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

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

cursor · 2026-10-05T08:14:27Z

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: 0b1d7966-9619-42f1-947d-f860cff471ae)

cursor · 2026-10-05T09:19:59Z

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: 72538497-e5fe-4ba2-ba8a-e1e968ae428a)

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 6e8d25cf8b

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

…escan cost Wording only (codex review on #1331): - the ACCUMULATE half exists (GroupAddr::Local / Via fold into a K-slot sink; K can be the compact destination count). The missing half is binding/resolving ONE destination from several coordinates in Report's CoordSpec / binder -- not an ndarray fold. - the sparse path costs one discovery scan per ordinal partition plus one run per fold state per tuple of the product of observed members, not one rescan per observed member. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01DCEP2fZdHYdMCtcEVcTpS2

cursor · 2026-10-05T10:21:43Z

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: add716df-d78d-4f56-b938-1c7c8b5bbfeb)

…r::Pair lance-graph-report planned a two-dimensional report as one population pass per partition member and called the composite-key group fold a named substrate gap. Quack already has GroupAddr::Pair, so the physical planner now picks a fold major: the widest other ordinal whose product with the fold key's domain fits domain_buffer_budget. Both coordinates then fold in ONE pass (20x30: 20 passes -> 1, 80 population scans -> 4). - Buckets, mask sets and a third ordinal stay partitions (no value-derived or multi-plane group key in Quack). - A pair-keyed SUM lowers to the NULL-preserving GroupReduce{SumI32}; its raw sink goes through Quack's normalize_group_sink (empty -> 0). Count/Min/Max already seed the report's own identities. - PhysicalPlan gains fold_major; explain labels it. Tests: new pair == per-member test over a mostly-empty space (cells, totals, hidden-axis merges; budget threshold 600 pairs / 599 does not); a disable run with the normalization removed goes red. t9 pins the pair programs plus a per-member twin; the pass-budget test moves to three ordinals. lance-graph-report 33/33, report-ogar 4/4, z8run-lance 10/10 (pivot_flow handle sizes and plan equality unchanged), clippy -D warnings clean. Board: one entry with the stack-convergence recon and the five decisions that scoped this cut. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01DCEP2fZdHYdMCtcEVcTpS2

…dy pair choice as OPEN Review nit: the knob now also bounds the product of the fold key's and the fold major's domains, not only a single coordinate's domain. Doc-only. The entry gains the review's non-blocking finding: the pair choice is greedy (widest key first), so 100x60x60 at budget 3600 runs 3600 passes where a joint (key, major) choice would run 100. REVISIT WHEN the planner is next touched; kept out of this PR to keep the fold key unchanged. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01DCEP2fZdHYdMCtcEVcTpS2

…n-resolving keyed fold) Correction before merge. GroupAddr::Pair addresses the dense product of two dimension domains; lowering every two-ordinal report to it made the Cartesian product the execution geometry because the primitive existed, not because a dense cube was demanded. The intended fold model resolves each row's accumulator destination directly, with accumulator state scaling with observed destinations (PowerShell hashtable / Excel pivot), not with |A| x |B|. - Reverted fold_major, fold_addr, Pair SUM normalization, the product budget, Pair strides/decode and the Pair-pinning tests: Report's planner is main's again. - exec.rs's gap doc now names the real missing primitive: a destination-resolving keyed fold (resolved key tuple -> compact slot in the same pass), substrate-first in ndarray's keyed-reduction family. Pair stays in Quack/mask-risc/ndarray for explicitly dense problems. - The board entry keeps the Pair cut as a superseded record and states why; decision 1 is struck; decisions 2-5 stand. lance-graph-report 32/32, fmt, clippy -D warnings clean. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01DCEP2fZdHYdMCtcEVcTpS2

… the greedy-Pair OPEN item Wording only. exec.rs gap doc: bind/resolve -> destination -> fold, with the law that a join/pivot used only to find an aggregate destination compiles to destination resolution + fold, never to an intermediate relation and never automatically to a dense product; the normal path binds destinations beforehand (ordinal, CAM/codebook, Via / composed reference) and only an unbindable universe uses a query-local compact resolver. No physical representation is chosen. Board entry: heading names the superseded Pair experiment; the greedy-Pair optimization leaves OPEN and is kept only as a historical observation under the superseded cut; the missing-primitive OPEN line no longer prescribes keys-SoA / observed-destination storage. Decisions 2-5 unchanged. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01DCEP2fZdHYdMCtcEVcTpS2

…escan cost Wording only (codex review on #1331): - the ACCUMULATE half exists (GroupAddr::Local / Via fold into a K-slot sink; K can be the compact destination count). The missing half is binding/resolving ONE destination from several coordinates in Report's CoordSpec / binder -- not an ndarray fold. - the sparse path costs one discovery scan per ordinal partition plus one run per fold state per tuple of the product of observed members, not one rescan per observed member. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01DCEP2fZdHYdMCtcEVcTpS2

… cost Wording only (codex review on #1331, operator-confirmed): - the missing capability is destination binding / coordinate resolution for Report. Once a compact destination ordinal or functional reference is bound, the existing GroupAddr::Local / Via keyed fold is the execution primitive. CoordSpec (Field, Bucket, MaskSet) cannot express a functional reference, a composed route, or several coordinates bound to one compact destination ordinal. A query-local interner is only a fallback. - Layout::Sparse costs one discovery scan per ordinal partition dimension, then T x FoldStates population folds, where T is the product of the partition radices (discovered members for ordinals, full radix for Bucket / MaskSet), plus a ReusedMask selection scan when chosen. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01DCEP2fZdHYdMCtcEVcTpS2

…te index Wording only, after rebasing onto main (#1331, #1333): - entries/README.md regenerated from the rebased tree (218 rows). - Fold-contract OPEN splits the old "destination resolution" item into A. destination binding / coordinate resolution (frontend / binder; lowers to the existing Local / Via fold; not a missing ndarray primitive) and B. functional route composition beyond depth 2 (substrate / R2IL; chain vs precomposed lane unmeasured). - TD-KEYED-SINK-MERGE-IDENTITY-1 states exactly what merge_group_sink checks (supported keyed i64 terminal, equal length) and adds binding identity, contribution-population identity, and the working law that an ordinal has meaning only inside its destination-space / codebook identity. - merge_group_sink precondition 5 names the binding. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01DCEP2fZdHYdMCtcEVcTpS2

AdaWorldAPI changed the title ~~report: fold two ordinal coordinates in one pass via Quack's GroupAddr::Pair~~ report: name the real fold gap (destination-resolving keyed fold); record why GroupAddr::Pair is not the pivot model Oct 5, 2026

AdaWorldAPI marked this pull request as ready for review October 5, 2026 10:15

AdaWorldAPI mentioned this pull request Oct 5, 2026

mask-risc: keyed partial-sink merge — MERGE made real (fold-contract Step E) #1332

Merged

chatgpt-codex-connector Bot reviewed Oct 5, 2026

View reviewed changes

Comment thread crates/lance-graph-report/src/exec.rs Outdated

Comment thread .claude/board/entries/2026-10-05-report-pair-key-and-stack-recon.md Outdated

claude added 5 commits October 5, 2026 10:22

AdaWorldAPI force-pushed the claude/report-groupaddr-pair branch from 9fd9292 to eb71675 Compare October 5, 2026 10:22

AdaWorldAPI changed the title ~~report: name the real fold gap (destination-resolving keyed fold); record why GroupAddr::Pair is not the pivot model~~ report: name the real gap (destination binding for Report); record why GroupAddr::Pair is not the pivot model Oct 5, 2026

AdaWorldAPI merged commit 4778f86 into main Oct 5, 2026
9 checks passed

Navigation Menu

Repository navigation

report: name the real gap (destination binding for Report); record why GroupAddr::Pair is not the pivot model - #1331

report: name the real gap (destination binding for Report); record why GroupAddr::Pair is not the pivot model#1331
AdaWorldAPI merged 6 commits into
mainfrom
claude/report-groupaddr-pair

AdaWorldAPI commented Oct 5, 2026 •

edited

Loading

Uh oh!

coderabbitai Bot commented Oct 5, 2026 •

edited

Loading

Uh oh!

cursor Bot commented Oct 5, 2026

Uh oh!

cursor Bot commented Oct 5, 2026

Uh oh!

chatgpt-codex-connector Bot left a comment

Uh oh!

Uh oh!

Uh oh!

cursor Bot commented Oct 5, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

AdaWorldAPI commented Oct 5, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Why the Pair path was wrong

The missing capability: destination binding for Report

What the source shows (recon, VERIFIED-IN-CODE)

This PR

Verification

Uh oh!

coderabbitai Bot commented Oct 5, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

cursor Bot commented Oct 5, 2026

Bugbot couldn't run - usage limit reached

Uh oh!

cursor Bot commented Oct 5, 2026

Bugbot couldn't run - usage limit reached

Uh oh!

chatgpt-codex-connector Bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

Uh oh!

Uh oh!

cursor Bot commented Oct 5, 2026

Bugbot couldn't run - usage limit reached

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

AdaWorldAPI commented Oct 5, 2026 •

edited

Loading

coderabbitai Bot commented Oct 5, 2026 •

edited

Loading