app: added a validation benchmark for lmm

app/lmm_benchmark.ml

+open Codepi
+
+let benchmark d =
+  let open Simulation_dataset in
+  Utils.recall_precision_curve
+    ~oracle:(oracle d)
+    ~labels:[|"topological";"multinomial";"gemma_score";"tdg09"|]
+    ~results:[
+      topological d ;
+      multinomial_asymptotic_lrt d ;
+      gemma ~lmm_test:`Score ~relatedness_mode:`Standardized d ;
+      tdg09 d ;
+    ]
+
+let () =
+  let sim =
+    let open Simulation_dataset in
+    convdet_simulation
+      ~tree:(NHX "example/trees4analyses_db/cyp_coding.Chrysithr_root.nhx")
+      ~profiles:"example/aa_fitness/263SelectedProfiles.tsv"
+      ~n_h0:100 ~n_ha:100 ~ne_s:(2., 2.) ()
+  in
+  let w = benchmark sim in
+  let loggers = [ Bistro_utils.Console_logger.create () ] in
+  Bistro_engine.Scheduler.simple_eval_exn ~loggers ~np:4 ~mem:(`GB 4) (Bistro.Workflow.path w)
+  |> print_endline