First revision of plotting script.

examples/CosmoVolume/plot_parallel_efficiency.py

+#!/usr/bin/env python
+#
+# Usage:
+#  python plot_parallel_efficiency input-file1-ext input-file2-ext ... index-into-threadListMax
+#
+# Description:
+# Plots speed up, parallel efficiency and time to solution given a stdout file generated by SWIFT.
+# 
+# Example:
+# python plot_parallel_efficiency _threads_cosma_stdout.txt _threads_knl_stdout.txt 8
+# 
+# The working directory should contain files 1_threads_cosma_stdout.txt - 64_threads_cosma_stdout.txt and 1_threads_knl_stdout.txt - 64_threads_knl_stdout.txt, i.e stdout for each run using a given number of threads
+
+import sys
+import numpy as np
+import matplotlib.pyplot as plt
+
+threadListMax=[1,2,4,8,16,24,32,64,128,192,256]
+skipToLine = (0,0,0) # Lines to skip before reading the wall clock times 
+version = ['Task Cleanup v0.3.0-247-g449de76a','Master v0.3.0-203-g6d9f54ae','KNL']
+linestyle = ('ro-','bo-','go-')
+cmdLine = './swift_fixdt -s -t 16 cosmoVolume.yml'
+kernel = 'Cubic Spline M4'
+platform = 'KNL'
+
+# Work out how many data series there are
+if len(sys.argv) == 3:
+  inputFileNames = (sys.argv[1],"")
+  maxNumThreads = int(sys.argv[2])
+  numOfSeries = 1
+elif len(sys.argv) == 4:
+  inputFileNames = (sys.argv[1],sys.argv[2])
+  maxNumThreads = int(sys.argv[3])
+  numOfSeries = 2
+elif len(sys.argv) == 5:
+  inputFileNames = (sys.argv[1],sys.argv[2],sys.argv[3])
+  maxNumThreads = int(sys.argv[4])
+  numOfSeries = 3
+
+threadList = threadListMax[0:maxNumThreads]
+
+# Parse file and return total time taken, speed up and parallel efficiency
+def parse_files():
+  
+  times = []
+  totalTime = []
+  serialTime = []
+  speedUp = []
+  parallelEff = []
+
+  for i in range(0,numOfSeries): # Loop over each version
+      
+      # Get the names of the branch and Git revision first
+      with open(str(threadList[0]) + inputFileNames[i], 'r') as f:
+          found_end = False
+          for line in f:
+            if 'Revision:' in line:
+                  s = line.split() 
+                  l = list(s)
+                  version[i] = s[3] + " " + s[1]
+                  version[i] = version[i][:-1]
+
+      times.append([])
+      totalTime.append([])
+      speedUp.append([])
+      parallelEff.append([])
+      for j in range(0,maxNumThreads): # Loop over no. of threads
+        times[i].append([])
+        file = str(threadList[j]) + inputFileNames[i]
+        times[i][j].append(np.loadtxt(file,skiprows=skipToLine[i],usecols=(5,)))
+        totalTime[i].append(np.sum(times[i][j]))
+      
+      serialTime.append(totalTime[i][0])
+      
+      for j in range(0,maxNumThreads): # Loop over no. of threads
+        speedUp[i].append(serialTime[i] / totalTime[i][j])
+        parallelEff[i].append(speedUp[i][j] / threadList[j])
+  
+  return (times,totalTime,speedUp,parallelEff)
+
+def print_results(times,totalTime,parallelEff,version):
+ 
+  for i in range(0,numOfSeries):
+    print " "
+    print "------------------------------------"
+    print version[i]
+    print "------------------------------------"
+    print "Wall clock time for: {} time steps".format(len(times[0][0][0]))
+    print "------------------------------------"
+    
+    for j in range(0,maxNumThreads):
+      print str(threadList[j]) + " threads: {}".format(totalTime[i][j])
+    
+    print " "
+    print "------------------------------------"
+    print "Parallel Efficiency for: {} time steps".format(len(times[0][0][0]))
+    print "------------------------------------"
+    
+    for j in range(0,maxNumThreads):
+      print str(threadList[j]) + " threads: {}".format(parallelEff[i][j])
+
+  return
+
+def plot_results(times,totalTime,speedUp,parallelEff):
+  
+  fig, axarr = plt.subplots(2, 2)
+  speedUpPlot = axarr[0, 0]
+  parallelEffPlot = axarr[0, 1]
+  totalTimePlot = axarr[1, 0]
+  emptyPlot = axarr[1, 1]
+  
+  # Plot speed up
+  for i in range(0,numOfSeries):
+    speedUpPlot.plot(threadList,speedUp[i],linestyle[i],label=version[i])
+  
+  speedUpPlot.plot(threadList,threadList,color='k',linestyle='--')
+  speedUpPlot.set_ylabel("Speed Up")
+  speedUpPlot.set_xlabel("No. of Threads")
+
+  # Plot parallel efficiency
+  for i in range(0,numOfSeries):
+    parallelEffPlot.plot(threadList,parallelEff[i],linestyle[i])
+  
+  parallelEffPlot.set_xscale('log')
+  parallelEffPlot.set_ylabel("Parallel Efficiency")
+  parallelEffPlot.set_xlabel("No. of Threads")
+  parallelEffPlot.set_ylim([0,1.1])
+
+  # Plot time to solution     
+  for i in range(0,numOfSeries):
+    totalTimePlot.loglog(threadList,totalTime[i],linestyle[i],label=version[i])
+  
+  totalTimePlot.set_xscale('log')
+  totalTimePlot.set_xlabel("No. of Threads")
+  totalTimePlot.set_ylabel("Time to Solution (ms)")
+  
+  totalTimePlot.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)
+  emptyPlot.axis('off')
+  
+  for i, txt in enumerate(threadList):
+    speedUpPlot.annotate(txt, (threadList[i],speedUp[0][i]))
+    parallelEffPlot.annotate(txt, (threadList[i],parallelEff[0][i]))
+    totalTimePlot.annotate(txt, (threadList[i],totalTime[0][i]))
+
+  fig.suptitle("Thread Speed Up, Parallel Efficiency and Time To Solution for {} Time Steps of Cosmo Volume\n Cmd Line: {}, Kernel: {}, Platform: {}".format(len(times[0][0][0]),cmdLine,kernel,platform))
+
+  return
+
+# Calculate results
+(times,totalTime,speedUp,parallelEff) = parse_files()
+
+plot_results(times,totalTime,speedUp,parallelEff)
+
+print_results(times,totalTime,parallelEff,version)
+
+plt.show()