[patnr]

Adds rows for the iterative ensemble smoothers LM-EnRML and GN-EnRML to the MiniRes benchmark, each with 4 iterations (as many updates as ES-MDA's 4 steps), and otherwise PET's defaults. The truncation is set in their iteration block, which is where they read it, to the 99% of the other rows.

RESULTS.md is not rerun here.

Tentative numbers, from run.py --quick (coarse case, 5 truths; means):

	misfit	CRPS forecast WWCT	logK RMSE	logK CRPS	sor CRPS	simulations
prior	97.4	0.103	0.772	0.437	0.0178
ES (one update)	14.4	0.060	0.856	0.494	0.0203	100
LM-EnRML (4 iterations)	1.82	0.027	0.745	0.425	0.0161	250
GN-EnRML (4 iterations)	5.53	0.025	0.779	0.440	0.0160	250
ES-MDA (4 steps, α = 4)	1.56	0.039	0.795	0.464	0.0269	250

Both accepted every step they tried: 250 simulations, i.e. the 50 members for the prior and for each of the 4 iterations.

Checked locally: the benchmark's smoke test, and ruff.

🤖 Generated with Claude Code