analyze.py

#!/usr/bin/env python
import csv
import difflib
import sys
import json

traces = {}
for row in csv.DictReader(open('SC_Android_Crash_Log.csv', 'r')):
  trace = row['stack_trace']
  if trace:
    if trace not in traces:
      traces[trace] = 0
    else:
      traces[trace] += 1

print "%d unique traces" % len(traces.keys())

groups  = []
for trace in traces.keys():
  if trace not in traces.keys(): continue
  sys.stderr.write('.')
  del traces[trace]
  groups.append([trace])
  for match in difflib.get_close_matches(trace, traces.keys(), n=len(traces)+1, cutoff=0.8):
    if match in traces: del traces[match]
    groups[-1].append(match)

print "\n%d similar traces" % len(groups)

out = open('groups.json', 'w')
json.dump(groups, out, sort_keys=True, indent=4)
out.close()