Pipeline now passes seed to add_indel script (currently a random seed per fna; to be fixed).

lib/dataset.ml

@@ -25,10 +25,10 @@ let repo ~preview dataset_l =
     )
   |> List.concat
 
-let add_indels_to_dataset d =
+let add_indels_to_dataset d ~seed =
     let p =  0.33 in
     let model_prefix = sprintf "%s_0.33_i" d.model_prefix in
     let tree_prefix  = d.tree_prefix in
     let is_real      = d.is_real in
-    let dataset      = Ready_dataset.add_indels_to_ready_dataset ~p d.dataset in
+    let dataset      = Ready_dataset.add_indels_to_ready_dataset ~p ~seed d.dataset in
     {model_prefix; tree_prefix; is_real; dataset; seed = Hashtbl.hash d.seed}

lib/pipeline.ml

@@ -132,8 +132,8 @@ let derive_from_tree ~tree_dir ~tree ~profile ~preview ~use_concat ~ns ~no_Ne ~n
   let dataset_H0_NeG5   = derive_from_model ~model:H0_NeG5   ~input_tree ~tree_dataset ~tree_prefix ~profile ~preview ~ns ~seed in
   let dataset_HaPCOC    = derive_from_model ~model:HaPCOC    ~input_tree ~tree_dataset ~tree_prefix ~profile ~preview ~ns ~seed in
   let dataset_HaPC_NeG5 = derive_from_model ~model:HaPC_NeG5 ~input_tree ~tree_dataset ~tree_prefix ~profile ~preview ~ns ~seed in
-  let indel_H0_NeG5 = Dataset.add_indels_to_dataset dataset_H0_NeG5 in
-  let indel_HaPC_NeG5 = Dataset.add_indels_to_dataset dataset_HaPC_NeG5 in
+  let indel_H0_NeG5 = Dataset.add_indels_to_dataset dataset_H0_NeG5 ~seed:(Random.int Int.max_value) in
+  let indel_HaPC_NeG5 = Dataset.add_indels_to_dataset dataset_HaPC_NeG5 ~seed:(Random.int Int.max_value) in
   let dataset_basis_hyps = [dataset_H0_NeG5; dataset_HaPCOC; dataset_HaPC_NeG5] in
   let models = Convergence_hypothesis.[
       [

lib/ready_dataset.ml

@@ -68,24 +68,25 @@ let paste d1 d2 =
   let fna_infos = Some (paste_fna_infos ~fna_infos_l) in
   let ready_dataset = of_raw {Raw_dataset.input_tree=r_d1.input_tree ; fna; fna_infos} in
   ready_dataset
-  
-  
-let add_indels_to_fna ~(p:float) (fna:nucleotide_fasta workflow) : nucleotide_fasta workflow =
-    let env = docker_image ~account:"carinerey" ~name:"python_basics" ~tag:"07252018" () in
-    workflow ~descr:("add_indels") [
+
+
+let add_indels_to_fna ~(p:float) ~(seed:int) (fna:nucleotide_fasta workflow) : nucleotide_fasta workflow =
+  let env = docker_image ~account:"carinerey" ~name:"python_basics" ~tag:"07252018" () in
+  workflow ~descr:("add_indels") [
     cmd "python" ~env [
       file_dump (string Scripts.add_indels) ;
       opt "-p" float p;
       opt "-a" dep fna;
       opt "-o" ident dest;
+      opt "-r" int seed;
       string "-c";
     ]
   ]
 
-let add_indels_to_ready_dataset ~p d =
+let add_indels_to_ready_dataset ~p ~seed d =
   let r_d = to_raw d in
-  let fna = add_indels_to_fna ~p r_d.fna in
+  (* VL: one seed per fna. FIXME:should be computed from global seed and task info instead *)
+  let fna = add_indels_to_fna ~p ~seed:(Random.int Int.max_value) r_d.fna in
   let fna_infos = r_d.fna_infos in
   let ready_dataset = of_raw {Raw_dataset.input_tree=r_d.input_tree ; fna; fna_infos} in
   ready_dataset
-