From 1aa9682787902cfbc1b7b8a5d88fe9ce7b7ac96e Mon Sep 17 00:00:00 2001
From: George Marchment <georgemarchment@yahoo.fr>
Date: Wed, 7 May 2025 16:54:30 +0200
Subject: [PATCH] Fix small bug
---
src/workflow.py | 44 +++++++++++++++++++++++++++-----------------
1 file changed, 27 insertions(+), 17 deletions(-)
diff --git a/src/workflow.py b/src/workflow.py
index bb2575f..30433b1 100644
--- a/src/workflow.py
+++ b/src/workflow.py
@@ -567,7 +567,8 @@ George Marchment, Bryan Brancotte, Marie Schmit, Frédéric Lemoine, Sarah Cohen
concordance_factor = 1,
uniformity_factor = 1,
min_nb_clusters_factor = 1,
- min_nb_non_relevant_cluster_factor = 1):
+ min_nb_non_relevant_cluster_factor = 1,
+ relevant_processes = []):#This parameter is to force relevant proceses is the user absolutely wants the
import copy
min_score, min_processes = np.inf, []
@@ -581,9 +582,9 @@ George Marchment, Bryan Brancotte, Marie Schmit, Frédéric Lemoine, Sarah Cohen
all_process_as_relevant.append(p.get_alias())
all_process_as_relevant = list(set(all_process_as_relevant))
#working_workflow.rewrite_workflow_remove_subworkflows(relevant_processes = all_process_as_relevant, render_graphs = False)
- #w_save = copy.deepcopy(working_workflow)
- w_save = copy.deepcopy(self)
+
w = copy.deepcopy(self)
+ #w = copy.deepcopy(working_workflow)
scripts_2_tools = {}
print("Extracting the tools from the processes")
print('-'*len(processes_called)+">")
@@ -601,20 +602,29 @@ George Marchment, Bryan Brancotte, Marie Schmit, Frédéric Lemoine, Sarah Cohen
#print(i/number_of_tries*100)
#w = copy.deepcopy(w_save)
- if(process_pre_selection == "bioinfo"):
- random_relevant_processes = w.get_random_relevant_processes_which_use_bioinformatics_tools(scripts_2_tools = scripts_2_tools)
- elif(process_pre_selection == "bioinfo_freq"):
- random_relevant_processes = w.get_random_relevant_processes_which_use_bioinformatics_tools_considering_their_frequency(scripts_2_tools = scripts_2_tools)
- elif(process_pre_selection == "None"):
- random_relevant_processes = w.generate_random_relevant_processes()
- else:
- raise Exception('process_pre_selection option not recognised')
- escape = 0
- while(escape<100 and set(random_relevant_processes) in already_tried):
+ def get_randomn_processes():
+ if(process_pre_selection == "bioinfo"):
+ random_relevant_processes = w.get_random_relevant_processes_which_use_bioinformatics_tools(scripts_2_tools = scripts_2_tools)
+ elif(process_pre_selection == "bioinfo_freq"):
+ random_relevant_processes = w.get_random_relevant_processes_which_use_bioinformatics_tools_considering_their_frequency(scripts_2_tools = scripts_2_tools)
+ elif(process_pre_selection == "None"):
+ random_relevant_processes = w.generate_random_relevant_processes()
+ else:
+ raise Exception('process_pre_selection option not recognised')
+ return list(set(random_relevant_processes+relevant_processes))
+
+ max_number_clusters = 30
+
+ random_relevant_processes = get_randomn_processes()
+ escape, escape_upper_bound = 0, 1000
+ while(escape<escape_upper_bound and (set(random_relevant_processes) in already_tried
+ or not(len(random_relevant_processes)<=np.min((reduction_beta*number_processes_called, max_number_clusters)))
+ or not(len(random_relevant_processes)>=reduction_alpha*number_processes_called))):
+ #print("here", random_relevant_processes)
escape+=1
- random_relevant_processes = w.generate_random_relevant_processes()
+ random_relevant_processes = get_randomn_processes()
#Cause it means we've already searched the majority of the possibilities
- if(escape>=100):
+ if(escape>=escape_upper_bound):
return min_processes
already_tried.append(set(random_relevant_processes))
#Here the nb of conditions returned is the number of conditions in the clusters after the rewrite
@@ -712,10 +722,10 @@ George Marchment, Bryan Brancotte, Marie Schmit, Frédéric Lemoine, Sarah Cohen
return score, cluster_with_processes, dico_results
- score, cluster_organisation, dico_results = get_score_from_set_relevant_processes(w_save, random_relevant_processes)
+ score, cluster_organisation, dico_results = get_score_from_set_relevant_processes(w, random_relevant_processes)
#print(dico_results)
if(len(cluster_organisation)>=reduction_alpha*number_processes_called and
- len(cluster_organisation)<=reduction_beta*number_processes_called and
+ len(cluster_organisation)<=np.min((reduction_beta*number_processes_called, max_number_clusters)) and
score<min_score):
#print()
#print("concordance", np.max(np.array(tab_nb_conditions_per_cluster)/np.array(tab_nb_executors_per_cluster)) )
--
GitLab