add script to create numpy array files from the statistics trace and separate the plot statistics functionality

scripts/parse_stats_trace.py

@@ -0,0 +1,26 @@
+#!/usr/bin/env python
+
+import argparse
+import numpy as np
+
+def read_stats_trace(stats_text_file):
+    allocated_regexp = r"dmmlib - ms all (\d+)"
+    allocated = np.fromregex(stats_text_file.name, allocated_regexp,
+            [('num', np.int64)])
+    requested_regexp = r"dmmlib - ms req (\d+)"
+    requested = np.fromregex(stats_text_file.name, requested_regexp,
+            [('num', np.int64)])
+    if allocated.size != requested.size:
+        print "Warning: Allocated and requested arrays have different sizes"
+    return (allocated['num'], requested['num'])
+
+if __name__=='__main__':
+    parser = argparse.ArgumentParser(
+            description="parse dmmlib's memory traces")
+    parser.add_argument("memory_trace", type=argparse.FileType('r'),
+                    help="dmmlib memory trace file")
+    args = parser.parse_args()
+    (allocated, requested) = read_stats_trace(args.memory_trace)
+    np.savez_compressed(args.memory_trace.name,
+            allocated=allocated, requested=requested)
+    args.memory_trace.close()

scripts/plot_memory_usage.py

@@ -1,4 +1,4 @@
-#!/usr/bin/python
+#!/usr/bin/env python
 
 import os, sys
 import subprocess

scripts/plot_stats.py

@@ -0,0 +1,66 @@
+#!/usr/bin/env python
+
+import argparse
+import numpy as np
+
+import matplotlib
+matplotlib.use('Agg')
+import matplotlib.pyplot as plt
+from matplotlib.ticker import FuncFormatter
+
+def megabytes(x, pos):
+    return '%1.f' % (x/(1024*1024))
+
+def get_memory_utilization_figure(allocated, requested):
+    memory_fig = plt.figure()
+    memory_ax = memory_fig.add_subplot(111)
+
+    memory_ax.plot(allocated, label='allocated memory', rasterized=True)
+    memory_ax.plot(requested, label='requested memory', rasterized=True)
+
+    memory_ax.set_ylim(ymin=0)
+
+    memory_ax.set_xticks([])
+    memory_ax.set_xlabel('Application timeline')
+
+    memory_ax.yaxis.set_major_formatter(FuncFormatter(megabytes))
+    memory_ax.get_yaxis().tick_left()
+
+    return memory_fig
+
+def get_memory_utilization_index(allocated, requested):
+    memory_fig = plt.figure()
+    memory_ax = memory_fig.add_subplot(111)
+
+    memory_ax.plot(np.true_divide(requested, allocated), rasterized=True)
+
+    memory_ax.set_xticks([])
+    memory_ax.set_xlabel('Application timeline')
+
+    memory_ax.get_yaxis().tick_left()
+    memory_ax.set_ylabel('Memory utilization index')
+
+    return memory_fig
+
+if __name__=='__main__':
+    parser = argparse.ArgumentParser(
+            description="parse dmmlib's memory traces")
+    parser.add_argument("stat_arrays_file", type=argparse.FileType('r'),
+                    help="compressed statistics file")
+    args = parser.parse_args()
+    npzfile = np.load(args.stat_arrays_file)
+    allocated = npzfile['allocated']
+    requested = npzfile['requested']
+    memory_fig = get_memory_utilization_figure(allocated, requested)
+    memory_fig.savefig(
+            '{}_mem_usage.pdf'
+            .format(args.stat_arrays_file.name.split('.')[0]),
+            dpi=150
+            )
+    memory_fig = get_memory_utilization_index(allocated, requested)
+    memory_fig.savefig(
+            '{}_mem_utilization_index.pdf'
+            .format(args.stat_arrays_file.name.split('.')[0]),
+            dpi=150
+            )
+    args.stat_arrays_file.close()