Network · Network-Adaptive Efficiency (NAMH, async) · async / long-running
namh_pipeline
NAMH Full Pipeline (async; seeded IAAFT; published namh). End-to-end published NAMH pipeline (namh::run_namh_pipeline v0.1.0, Bhandari & Sahu 2026): rolling DFA-Hurst → KSG TE (Euclidean RANN) → IAAFT effective TE → Benjamini-Hochberg FDR adjacency → Hurst-weighted TE → NAMH fixed point → Leiden communities → centralities. RNG honesty: the package's IAAFT generator is unseeded, so this runner pins RNGkind L'Ecuyer-CMRG + set.seed(seed, default 42) — results are reproducible for a fixed {seed, n_cores} (verified: same seed twice → identical surrogate p-values; different seed → different). Gate honesty (D1): the paper's own FDR gate is the only gate exposed — at the canonical config it retains 0/552 directed edges in every window, so downstream network surfaces are degenerate by construction and reported as measured. ~552 pairs × (1+n_surrogates) KSG estimates per window (≈24 min at B=200) — async-only via /api/jobs/submit; rejected by the sync /api/compute/run.
Identity
version 1.0.0 · capability network · min_obs 252 · runner r (namh_pipeline.R)
Primitives
P1P3P4P5P6
Composes with
not yet a catalog field — operator-composition metadata lands with Pathway F; nothing is invented here.
Parameters
| name | type | req | values / range |
|---|
| window | int | optional | |
| step | int | optional | |
| k_nn | int | optional | |
| n_surrogates | int | optional | |
| fdr_alpha | num | optional | |
| lambda | num | optional | |
| seed | int | optional | |
| n_cores | int | optional | |
| window_index | int | optional | |
Returns
method,dataset,config,n_series,runtime_s,per_window,note,source
Paper
none — honest: a substrate method; no single companion paper claims it.
Changelog
[object Object]
Run it
curl -s -X POST https://shssm-compute-b7ui3oxaqq-el.a.run.app/api/compute/run \
-H "Content-Type: application/json" \
-d '{"method":"namh_pipeline","params":{}}'