Merge pull request #45 from martinghunt/one_thread_not_use_multithreaded

One thread not use multithreaded

ariba/assembly_variants.py

@@ -1,3 +1,4 @@
+import sys
 import operator
 import pyfastaq
 import pymummer
@@ -305,7 +306,7 @@ def get_variants(self, ref_sequence_name, nucmer_coords):
 
                     # include new variant, except if the ref type is variants only and
                     # the new variant matches to a known variant
-                    if new_variant is not None and (ref_sequence_type != 'variants_only' or len(new_variant[5]) > 0):
+                    if new_variant is not None and (ref_sequence_type != 'variants_only' or len(new_variant[5]) > 0 or new_variant[3] == 'INDELS'):
                             variants[contig].append(new_variant)
                     used_known_variants.update(used_variants)
 

ariba/cluster.py

@@ -120,15 +120,15 @@ def _count_reads(reads1, reads2):
 
     def _clean(self):
         if self.clean == 0:
-            print('   ... not deleting anything because --clean 0 used', file=self.log_fh)
+            print('   ... not deleting anything because --clean 0 used', file=self.log_fh, flush=True)
             return
         elif self.clean == 2:
             print('    rm -r ', self.root_dir)
             shutil.rmtree(self.root_dir)
             return
 
         if os.path.exists(self.assembly_dir):
-            print('    rm -r', self.assembly_dir, file=self.log_fh)
+            print('    rm -r', self.assembly_dir, file=self.log_fh, flush=True)
             shutil.rmtree(self.assembly_dir)
 
         to_delete = [
@@ -146,7 +146,7 @@ def _clean(self):
 
         for filename in to_delete:
             if os.path.exists(filename):
-                print('    rm', filename, file=self.log_fh)
+                print('    rm', filename, file=self.log_fh, flush=True)
                 try:
                     os.unlink(filename)
                 except:
@@ -157,7 +157,7 @@ def run(self):
         self.logfile = os.path.join(self.root_dir, 'log.txt')
         self.log_fh = pyfastaq.utils.open_file_write(self.logfile)
 
-        print('Choosing best reference sequence:', file=self.log_fh)
+        print('Choosing best reference sequence:', file=self.log_fh, flush=True)
         seq_chooser = best_seq_chooser.BestSeqChooser(
             self.reads1,
             self.reads2,
@@ -174,7 +174,7 @@ def run(self):
             self.status_flag.add('ref_seq_choose_fail')
             self.assembled_ok = False
         else:
-            print('\nAssembling reads:', file=self.log_fh)
+            print('\nAssembling reads:', file=self.log_fh, flush=True)
             self.ref_sequence_type = self.refdata.sequence_type(self.ref_sequence.id)
             assert self.ref_sequence_type is not None
             self.assembly = assembly.Assembly(
@@ -199,7 +199,7 @@ def run(self):
             self.assembled_ok = self.assembly.assembled_ok
 
         if self.assembled_ok:
-            print('\nAssembly was successful\n\nMapping reads to assembly:', file=self.log_fh)
+            print('\nAssembly was successful\n\nMapping reads to assembly:', file=self.log_fh, flush=True)
 
             mapping.run_bowtie2(
                 self.reads1,
@@ -218,14 +218,14 @@ def run(self):
             if self.assembly.has_contigs_on_both_strands:
                 self.status_flag.add('hit_both_strands')
 
-            print('\nMaking and checking scaffold graph', file=self.log_fh)
+            print('\nMaking and checking scaffold graph', file=self.log_fh, flush=True)
 
             bam_parser = bam_parse.Parser(self.final_assembly_bam, self.assembly.sequences)
             bam_parser.parse()
             if not bam_parser.scaff_graph_is_consistent(self.min_scaff_depth, self.max_insert):
                 self.status_flag.add('scaffold_graph_bad')
 
-            print('Comparing assembly against reference sequence', file=self.log_fh)
+            print('Comparing assembly against reference sequence', file=self.log_fh, flush=True)
             self.assembly_compare = assembly_compare.AssemblyCompare(
               self.final_assembly_fa,
               self.assembly.sequences,
@@ -256,7 +256,7 @@ def run(self):
                     break
 
 
-            print('\nCalling variants with samtools:', file=self.log_fh)
+            print('\nCalling variants with samtools:', file=self.log_fh, flush=True)
 
             self.samtools_vars = samtools_variants.SamtoolsVariants(
                 self.final_assembly_fa,
@@ -274,14 +274,14 @@ def run(self):
             if self.samtools_vars.variants_in_coords(self.assembly_compare.assembly_match_coords(), self.samtools_vars.vcf_file):
                 self.status_flag.add('variants_suggest_collapsed_repeat')
         else:
-            print('\nAssembly failed\n', file=self.log_fh)
+            print('\nAssembly failed\n', file=self.log_fh, flush=True)
             self.status_flag.add('assembly_fail')
 
-        print('\nMaking report lines', file=self.log_fh)
+        print('\nMaking report lines', file=self.log_fh, flush=True)
         self.report_lines = report.report_lines(self)
-        print('\nCleaning with clean option ', self.clean, ':', sep='', file=self.log_fh)
+        print('\nCleaning with clean option ', self.clean, ':', sep='', file=self.log_fh, flush=True)
         self._clean()
-        print('Finished', file=self.log_fh)
+        print('Finished', file=self.log_fh, flush=True)
         pyfastaq.utils.close(self.log_fh)
 
         # This stops multiprocessing complaining with the error:

ariba/clusters.py

@@ -282,8 +282,13 @@ def _init_and_run_clusters(self):
                 ))
 
 
-        pool = multiprocessing.Pool(self.threads)
-        cluster_list = pool.starmap(_run_cluster, zip(cluster_list, itertools.repeat(self.verbose)))
+        if self.threads > 1:
+            pool = multiprocessing.Pool(self.threads)
+            cluster_list = pool.starmap(_run_cluster, zip(cluster_list, itertools.repeat(self.verbose)))
+        else:
+            for c in cluster_list:
+                _run_cluster(c, self.verbose)
+
         self.clusters = {c.name: c for c in cluster_list}