Added profiling

sharpy/sharpy_main.py

@@ -45,6 +45,15 @@ def main(args=None, sharpy_input_dict=None):
     import sharpy.postproc
     import sharpy.generators
     import sharpy.controllers
+
+    # Profiling
+    prof_enable = True
+
+    if prof_enable:
+        import cProfile, sharpy.utils.profile_analyse
+        pr = cProfile.Profile()
+        pr.enable()
+
     # ------------
 
     try:
@@ -148,6 +157,11 @@ def main(args=None, sharpy_input_dict=None):
         cout.cout_wrap('FINISHED - CPU process time = %f6 seconds' % cpu_time, 2)
         finish_writer()
 
+        # End profiling and analse output using string arguments
+        if prof_enable:
+            pr.disable()
+            prof_data = sharpy.utils.profile_analyse.profile_analse(pr, settings)
+
     except Exception as e:
         try:
             logdir = settings['SHARPy']['log_folder'] + '/' + settings['SHARPy']['case']
@@ -165,8 +179,7 @@ def main(args=None, sharpy_input_dict=None):
         logging.info('SHARPy Error Log')
         logging.error("Exception occurred", exc_info=True)
         raise e
-
-    return data
+    return data, prof_data
 
 
 

sharpy/utils/profile_analyse.py

@@ -0,0 +1,38 @@
+import string, csv, string, os, numpy as np, sys, io, cProfile, pstats
+from pstats import SortKey
+
+def profile_analse(pr, settings, names = ['beam', 'uvlm']):
+
+    prof_dir = settings['SHARPy']['route'] + settings['SHARPy']['case'] + '_profile.txt'
+    sys.stdout = open(prof_dir, 'w')
+    s = io.StringIO()
+    ps = pstats.Stats(pr, stream=s).sort_stats('tottime')            
+    ps.print_stats()
+    print(s.getvalue())
+    sys.stdout.close()
+    sys.stdout = sys.__stdout__
+    if names[-1] != "other":
+        names.append("other")
+    t_names = np.zeros(len(names), dtype=float)
+    t_total = 0.0
+
+    with open(prof_dir, newline='\n') as f: 
+            csv_f = csv.reader(f, delimiter=' ', skipinitialspace=True)
+
+            for row in csv_f:
+                try:
+                    t_total += float(list(row)[1])
+                    for i in range(len(names)-1):
+                        if list(row)[5].find(names[i]) != -1:
+                            t_names[i] += float(list(row)[1])
+                except: () 
+
+    t_names[-1] = t_total - np.sum(t_names)
+
+    print("\n")        
+    for i in range(len(names)):
+        print(names[i] + ' time = ' + f'{t_names[i]:.3f}' + ' seconds (' + f'{t_names[i]/t_total*100:.3f}' + '%)')
+    print('\nTotal time = ' + f'{t_total:.3f}' + ' seconds\n')
+
+
+    return {'names':names, 't_names':t_names, 't_total':t_total}