RP curve: allow to choose column in result file

app/lmm_benchmark.ml

@@ -4,23 +4,24 @@ let benchmark d =
   let open Simulation_dataset in
   Utils.recall_precision_curve
     ~oracle:(oracle d)
-    ~labels:[| "topological"; "multinomial";"gemma_score";"inhouse"; "tdg09" |]
+    ~labels:[| "topological"; "multinomial";"gemma_score";"inhouse"; "pcoc"; (* "pcocv2" ; *)"tdg09" |]
     ~results:[
-      topological d ;
-      multinomial_asymptotic_lrt d ;
-      gemma ~lmm_test:`Score ~relatedness_mode:`Standardized d ;
-      inhouse_lmm d ;
-      tdg09 d ;
+      topological d, 1 ;
+      multinomial_asymptotic_lrt d, 1 ;
+      gemma ~lmm_test:`Score ~relatedness_mode:`Standardized d, 1 ;
+      inhouse_lmm d, 1 ;
+      pcoc d, 3 ;
+      (* pcoc_v2 d ; *)
+      tdg09 d, 1 ;
     ]
 
 let () =
   let sim =
     let open Simulation_dataset in
     convdet_simulation
-      (* ~tree:(NHX "example/trees4analyses_db/cyp_coding.Chrysithr_root.nhx") *)
-      ~tree:(NHX "example/trees_test/orthomam_rooted.nhx")
+      ~tree:(NHX "data/besnard2009/besnard2009.nhx")
       ~profiles:"example/aa_fitness/263SelectedProfiles.tsv"
-      ~n_h0:900 ~n_ha:100 ~ne_s:(2., 2.) ()
+      ~n_h0:900 ~n_ha:100 ~branch_scale:3. ~ne_s:(2., 2.) ()
   in
   let w = benchmark sim in
   let loggers = [ Bistro_utils.Console_logger.create () ] in

lib/simulation_dataset.ml

@@ -164,10 +164,10 @@ let multinomial_benchmark d =
     ~oracle:(oracle d)
     ~labels:[|"LRT";"LRTsim";"sparse";"sparse_sim"|]
     ~results:[
-      multinomial_asymptotic_lrt d ;
-      multinomial_simulation_lrt d ;
-      multinomial_asymptotic_sparse d ;
-      multinomial_simulation_sparse d ;
+      multinomial_asymptotic_lrt d, 1 ;
+      multinomial_simulation_lrt d, 1 ;
+      multinomial_asymptotic_sparse d, 1 ;
+      multinomial_simulation_sparse d, 1 ;
     ]
 
 
@@ -199,11 +199,11 @@ let benchmark2 d =
     ~oracle:(oracle d)
     ~labels:[|"identical";"topological";"multinomial";"pcoc";"tdg09"|]
     ~results:[
-      identical d ;
-      topological d ;
-      multinomial_asymptotic_lrt d ;
-      pcoc d ;
-      tdg09 d ;
+      identical d, 1 ;
+      topological d, 1 ;
+      multinomial_asymptotic_lrt d, 1 ;
+      pcoc d, 3 ;
+      tdg09 d, 1 ;
     ]
 
 (* param exploration for SMBE paper *)

lib/utils.ml

@@ -37,11 +37,11 @@ let%pworkflow recall_precision_curve ~labels ~oracle ~results : pdf file =
   let open Codepitk in
   let n = Array.length labels in
   assert (n = List.length results) ;
-  let colors = Array.sub [|"blue";"green";"red";"tan1";"royalblue4";"darkorchid"|] ~pos:0 ~len:n in
+  let colors = Array.sub [|"blue";"green";"red";"tan1";"royalblue4";"darkorchid";"yellow"|] ~pos:0 ~len:n in
   let labels = [%param labels] in
-  let load_results fn =
+  let load_results fn col =
     let df = Dataframe.from_file fn  |> Rresult.R.failwith_error_msg in
-    match Dataframe.get_col df 1 with
+    match Dataframe.get_col df col with
     | Some (Floats xs) -> Array.map xs ~f:Option.some
     | Some (Float_opts xs) -> xs
     | Some (Ints xs) -> Array.map xs ~f:(fun i -> Some (Float.of_int i))
@@ -49,10 +49,11 @@ let%pworkflow recall_precision_curve ~labels ~oracle ~results : pdf file =
     | _ -> failwith "expected a numeric column at pos 1"
   in
   let oracle =
-    load_results [%path oracle]
+    load_results [%path oracle] 1
     |> Array.map ~f:(Option.map ~f:Float.(( <> ) 0.))
   in
-  let results = List.map ~f:load_results [%eval Bistro.Workflow.path_list results] in
+  let result_paths = [%eval Bistro.Workflow.path_list (List.map ~f:fst results)] in
+  let results = List.map2_exn result_paths (List.map ~f:snd results) ~f:load_results in
   let filter_table x y =
     Array.filter_mapi x ~f:(fun i x_i ->
         match x_i, y.(i) with