ADR 0010: pushdown math, per-encoding table, PoC findings#34
Open
dfa1 wants to merge 1 commit into
Open
Conversation
Adds the concrete math behind compute pushdown: encode the threshold
into the encoding's integer domain, integer-compare, decode only
matches; plus a boundary re-check for FP edge cases.
Per-encoding table identifies which encodings support pushdown
(ALP, FoR, Bitpacked, Dict — order-preserving invertible) and which
do not (ZigZag — order-non-preserving; Pco/Zstd/FSST — opaque blocks).
Adds PoC findings from worktree lazy-alp-f64-poc:
- lazy ALP is +10.6% on full fold (drops the hasFilter() gate)
- ALP-level pushdown is +100% at sel=0.1
- residual 2-3× JNI gap at low sel is Bitpacked SIMD, ADR 0005
territory
Drops the hasFilter() gate from the final design and notes the
supersession inline on the earlier sections so the reasoning chain
stays readable.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
lazy-alp-f64-pocworktree: lazy ALP is +10.6% on full fold, ALP-level pushdown is +100% at sel=0.1, residual JNI gap at low selectivity is Bitpacked SIMD (ADR 0005 territory)hasFilter()gate from the final design — lazy is strictly faster than eager on full fold, so the gate is unnecessaryTest plan
🤖 Generated with Claude Code