fix: use fork context instead of spawn to avoid forkserver crash and slowdown

Author: JoseCorCab

SigProfilerClusters/SigProfilerClusters.py

@@ -1225,7 +1225,7 @@ def analysis(
         original_samples_length = os.listdir(original_samples)
         if processors > len(original_samples_length):
             max_seed = len(original_samples_length)
-        pool = mp.Pool(max_seed)
+        pool = mp.get_context("fork").Pool(max_seed)
         # pool_break = len(simulations)/max_seed
         samples_parallel = [[] for i in range(max_seed)]
         pool_bin = 0
@@ -1259,7 +1259,7 @@ def analysis(
         # Parallelize and calculate distances for the simulated samples
         if processors > len(simulations):
             max_seed = len(simulations)
-        pool = mp.Pool(max_seed)
+        pool = mp.get_context("fork").Pool(max_seed)
         # sim_break = len(simulations)/max_seed
         simulations_parallel = [[] for i in range(max_seed)]
         sim_bin = 0

SigProfilerClusters/SigProfilerHotSpots.py

@@ -1029,7 +1029,7 @@ def analysis(
         original_samples_length = os.listdir(original_samples)
         if processors > len(original_samples_length):
             max_seed = len(original_samples_length)
-        pool = mp.Pool(max_seed)
+        pool = mp.get_context("fork").Pool(max_seed)
         # pool_break = len(simulations)/max_seed
         samples_parallel = [[] for i in range(max_seed)]
         pool_bin = 0
@@ -1063,7 +1063,7 @@ def analysis(
         # Parallelize and calculate distances for the simulated samples
         if processors > len(simulations):
             max_seed = len(simulations)
-        pool = mp.Pool(max_seed)
+        pool = mp.get_context("fork").Pool(max_seed)
         # sim_break = len(simulations)/max_seed
         simulations_parallel = [[] for i in range(max_seed)]
         sim_bin = 0

SigProfilerClusters/classifyFunctions.py

@@ -1713,7 +1713,7 @@ def findClustersOfClusters(
         max_seed = len(subclasses)
     else:
         max_seed = processors
-    pool = mp.Pool(max_seed)
+    pool = mp.get_context("fork").Pool(max_seed)
 
     subclasses_parallel = [[] for i in range(max_seed)]
     subclass_bin = 0
@@ -2620,7 +2620,7 @@ def findClustersOfClusters_noVAF(
         max_seed = len(subclasses)
     else:
         max_seed = processors
-    pool = mp.Pool(max_seed)
+    pool = mp.get_context("fork").Pool(max_seed)
 
     subclasses_parallel = [[] for i in range(max_seed)]
     subclass_bin = 0