Library coverage

All eleven adapters below compile and have passed 429 baseline correctness cases twice plus 66 targeted pivot/sign cases. The expanded JMH comparison reruns all adapters with identical harness settings. See completed results.

Distribution and backend distinctions

The original four-library report remains available as a separate historical run. Its omissions were integration scope decisions, not evidence that other libraries could not run.

Current version selection

The latest comparison upgrades the primary Commons adapter to Commons Math 4.0-SNAPSHOT, Parallel Colt to 0.11.4 from its upstream release tag, and NM Dev to the validated 1.2.6 release JAR. See the complete version audit for all eleven libraries, release-versus-snapshot distinctions and exact checksums. Version labels below are read from the selected run.

Measured backend records

Library Measured version / backend
NMDEV NM Dev 1.2.6 — Validated combined obfuscated 1.2.6 JAR; exact hash and CI source in validated-nmdev-artifact.json
EJML EJML 0.46.1 — EJML 0.46.1
COMMONS Commons Math 4.0-SNAPSHOT (20261018.085925-1575) — Apache Commons Math4 legacy linear algebra; pinned official development snapshot 4.0-20261018.085925-1575
JAMA JAMA 1.0.3 — JAMA 1.0.3
COLT Colt 1.2.0 — Colt 1.2.0
PARALLEL_COLT Parallel Colt 0.11.4 (upstream tag build) — Upstream 0.11.4 release tag compiled without algorithm modifications; ConcurrencyUtils threads=1
JBLAS jblas 1.2.5 (native) — Bundled native libjblas; requested native environment threads=1; actual native worker counts not independently audited
F2J_LAPACK F2J LAPACK/BLAS 0.1 — Direct translated Java BLAS/LAPACK distribution arpack_combined_all0.1; replacement distribution, not JLAPACK0.8
MTJ MTJ 1.0.4 (F2J) — F2jBLAS/F2jLAPACK/F2jARPACK explicitly forced by fork flags
OJALGO ojAlgo 57.5.0 — R064Store/decomposition factories; library thread limit=1; deferred scalar results materialized inside timing
UJMP UJMP 0.3.0 (core) — Core algorithms only; external backends disabled; threads=1