Benchmark
=========

Detailed evaluation (paper-style): :doc:`evaluation`.

Regenerate all artifacts
------------------------

.. code-block:: bash

   python scripts/run_all_benchmarks.py

Writes ``docs/source/_generated/*.json`` and ``*.rst`` for Sphinx.

Synthetic only (no ``tcl_test`` data):

.. code-block:: bash

   python scripts/generate_benchmark_report.py

Real data only (requires ``tcl_test`` + conda env ``st`` for R reference):

.. code-block:: bash

   python scripts/run_tcl_benchmark.py --also-docs

Real data (GSE288946)
---------------------

.. include:: _generated/tcl_benchmark_table.rst
   :start-line: 0

Synthetic scaling
-----------------

.. include:: _generated/benchmark_table.rst
   :start-line: 0

Tests
-----

.. code-block:: bash

   pytest tests/test_theta_od_parity.py -q
   pytest tests/test_parity_r.py -m parity_r   # optional, needs R