Add rev18 clutter-optimization branch with monte_carlo_clutter_rev5 and rev11-only validation/profiling

[nicklasorte]

Motivation

• Redo the clutter helper optimization in a correctness-first way while preserving subchunk_agg_check_maxazi_rev11 semantics as the golden behavioral baseline.
• Isolate the change to a fresh helper revision and a fresh subchunk revision so acceptance is measured only vs rev11 and not vs rev14/rev15.

Description

• Added monte_carlo_clutter_rev5_app.m, a vectorized, shape-safe replacement for monte_carlo_clutter_rev3_app that preserves the rev3 output contract, units, nearestpoint semantics, and NaN/Inf cleanup.
• Added subchunk_agg_check_maxazi_rev18.m, copied from rev11 semantics with the sole functional change of calling monte_carlo_clutter_rev5_app (rev3->rev5) and a header noting the rev11 validation requirement.
• Added validate_subchunk_agg_check_maxazi_rev11_rev18_statistical.m, which runs rev11 and rev18 on identical inputs, collects runtime, computes mean/std/min/max/median/p90/p95/p99, applies fail-closed core/tail thresholds (defaults: core_abs=0.250 dB, core_rel=0.020, tail_abs=0.150 dB, tail_rel=0.010), prints a structured PASS/FAIL summary, and returns a results struct.
• Added profile_subchunk_agg_check_maxazi_rev18_real.m, which profiles both rev11 and rev18 on identical inputs, builds a timing table for requested functions (including clutter helper, PR, EIRP helper, nearestpoint_app, db2pow), reports clutter helper drop fraction vs rev11, and returns a profiling results struct.

Testing

• Performed static sanity checks: verified the four new files exist and declared function signatures using rg/file listing and inspected headers.
• Committed the new files to the repository to record the change set.
• Attempted to run an execution environment check (octave --version), but execution failed because Octave/MATLAB is not available in this container (/bin/bash: octave: command not found), so runtime validation and profiling were not executed here.
• Note: the repository now contains the validation (validate_subchunk_agg_check_maxazi_rev11_rev18_statistical) and profiling (profile_subchunk_agg_check_maxazi_rev18_real) harnesses which must be run in a MATLAB/Octave environment to produce numeric PASS/FAIL and profiling results against real inputs.

---

Codex Task

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: c87197e2a2

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".

[chatgpt-codex-connector]

Preserve Inf handling parity with rev3 in clutter interpolation

subtract_Pr(~isfinite(subtract_Pr))=0 changes rev3 semantics when the interpolation endpoints include Inf. In rev3, an infinite temp_diff_Pr propagates into monte_carlo_clutter_loss and is then normalized to 0 by the final isinf cleanup; here the value is zeroed earlier, so the output becomes prev_loss instead of 0. This can materially change aggregate interference results on datasets with infinite clutter values and violates the stated rev11-equivalence contract.

Useful? React with 👍 / 👎.