From 178fc8c95adce6d3c572b67ddd703964ce18f644 Mon Sep 17 00:00:00 2001 From: Chris McComb Date: Tue, 13 Oct 2026 02:03:23 -0400 Subject: [PATCH] Document nonlinear OSWEC source HDF5 provenance --- tests/fixtures/matlab/README.md | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/tests/fixtures/matlab/README.md b/tests/fixtures/matlab/README.md index 59d56fc5..cf21091a 100644 --- a/tests/fixtures/matlab/README.md +++ b/tests/fixtures/matlab/README.md @@ -17,6 +17,13 @@ The archive contains `time`, `pose`, `velocity`, `acceleration`, `63f3669b192ab271a92f5ea9c4650d1d183ef68f0abc4a6b0faedcd2b4948e4f`. The refined-step source job gates its 0.1 s run against these published channels before evaluating step sensitivity or independent Python motion. +The motion archive does not contain hydrodynamic coefficients. The pinned +[refinement artifact](https://github.com/cmudrc/wec-sim-python/actions/runs/38093454463) +supplies its matching `oswec.h5` (SHA-256 +`72664d60037698350513d6481d4114db79fe2cf8984247a13ff16e980c52d31f`). +That HDF5 has 36 excitation headings; `examples/data/oswec.h5` has 10 and +also differs in added mass. Use the source-matched HDF5 for this motion +comparison rather than substituting the example file. ## RM3 regular wave