Iteratively increase sample size for good/bad classification.

scripts/slave/recipe_modules/auto_bisect/perf_revision_state.py

 # Copyright 2015 The Chromium Authors. All rights reserved.
 # Use of this source code is governed by a BSD-style license that can be
 # found in the LICENSE file.
 
 import json
+import math
 import tempfile
 import os
 import uuid
 
 from . import revision_state
 
 if 'CACHE_TEST_RESULTS' in os.environ:  # pragma: no cover
   from . import test_results_cache
 
+# These relate to how to increase the number of repetitions during re-test
+MINIMUM_SAMPLE_SIZE = 5
+INCREASE_FACTOR = 1.5
 
 class PerfRevisionState(revision_state.RevisionState):
   """Contains the state and results for one revision in a perf bisect job."""
 
   def __init__(self, *args, **kwargs):
     super(PerfRevisionState, self).__init__(*args, **kwargs)
     self.values = []
     self.mean_value = None
     self.std_dev = None
+    self.repeat_count = MINIMUM_SAMPLE_SIZE
     self._test_config = None
 
-  def _read_test_results(self):
+  def _read_test_results(self, check_revision_goodness=True):
     """Gets the test results from GS and checks if the rev is good or bad."""
     test_results = self._get_test_results()
     # Results will contain the keys 'results' and 'output' where output is the
     # stdout of the command, and 'results' is itself a dict with the key
     # 'values' unless the test failed, in which case 'results' will contain
     # the 'error' key explaining the type of error.
     results = test_results['results']
     if results.get('errors'):
       self.status = PerfRevisionState.FAILED
       if 'MISSING_METRIC' in results.get('errors'):  # pragma: no cover
         self.bisector.surface_result('MISSING_METRIC')
       return
-    self.values = results['values']
+    self.values = (self.values or []) + results['values']

[dtu, 2016/02/05 00:18:02]

Why do you need this? Isn't self.values already initialized to []?

     if self.bisector.is_return_code_mode():
       retcodes = test_results['retcodes']
       overall_return_code = 0 if all(v == 0 for v in retcodes) else 1
       self.mean_value = overall_return_code
     elif self.values:
       api = self.bisector.api
       self.mean_value = api.m.math_utils.mean(self.values)
       self.std_dev = api.m.math_utils.standard_deviation(self.values)
     # Values were not found, but the test did not otherwise fail.
     else:
       self.status = PerfRevisionState.FAILED
       self.bisector.surface_result('MISSING_METRIC')
       return
     # We cannot test the goodness of the initial rev range.
-    if self.bisector.good_rev != self and self.bisector.bad_rev != self:
+    if (self.bisector.good_rev != self and self.bisector.bad_rev != self and
+        check_revision_goodness):
       if self._check_revision_good():
         self.good = True
       else:
         self.bad = True
 
   def _write_deps_patch_file(self, build_name):
     """Saves the DEPS patch in a temp location and returns the file path."""
     api = self.bisector.api
     file_name = str(api.m.path['tmp_base'].join(build_name + '.diff'))
     api.m.file.write('Saving diff patch for ' + str(self.revision_string),
@@ skipping 41 matching lines @@
           api.m.file.remove('cleaning up patch', self.patch_file)
         except api.m.step.StepFailure:  # pragma: no cover
           print 'Could not clean up ' + self.patch_file
 
   def _get_bisect_config_for_tester(self):
     """Copies the key-value pairs required by a tester bot to a new dict."""
     result = {}
     required_test_properties = {
         'truncate_percent',
         'metric',
-        'max_time_minutes',
         'command',
-        'repeat_count',
         'test_type'
     }
     for k, v in self.bisector.bisect_config.iteritems():
       if k in required_test_properties:
         result[k] = v
+    result['repeat_count'] = self.repeat_count
     self._test_config = result
     return result
 
   def _do_test(self):
     """Triggers tests for a revision, either locally or via try job.
 
     If local testing is enabled (i.e. director/tester merged) then
     the test will be run on the same machine. Otherwise, this posts
     a request to buildbot to download and perf-test this build.
     """
@@ skipping 50 matching lines @@
       builder = self.bisector.get_builder_bot_for_this_platform()
     if self.status == PerfRevisionState.TESTING:
       builder = self.bisector.get_perf_tester_name()
     return {
         'type': 'buildbot',
         'master': master,
         'builder': builder,
         'job_name': self.job_name,
     }
 
+  def retest(self):  # pragma: no cover
+    # We need at least 5 samples for the hypothesis testing to work properly.

[dtu, 2016/02/05 00:18:02]

Sorry, not strictly true as written. Only true with these assumptions:
- Mann-Whitney U
- 2-sided
- Significance level of 0.01

+    target_sample_size =  max(5, math.ceil(len(self.values) * 1.5))
+    self.status = PerfRevisionState.NEED_MORE_DATA
+    self.repeat_count = target_sample_size - len(self.values)
+    self.start_job()
+    self.bisector.wait_for_any([self])
+
   def _get_test_results(self):
     """Tries to get the results of a test job from cloud storage."""
     api = self.bisector.api
     try:
       stdout = api.m.raw_io.output()
       name = 'Get test results for build ' + self.commit_hash
       step_result = api.m.gsutil.cat(self.test_results_url, stdout=stdout,
                                      name=name)
     except api.m.step.StepFailure:  # pragma: no cover
       self.bisector.surface_result('TEST_FAILURE')
       return None
     else:
       return json.loads(step_result.stdout)
 
   def _check_revision_good(self):
     """Determines if a revision is good or bad.
 
-    Note that our current approach is to determine whether it is closer to
-    either the 'good' and 'bad' revisions given for the bisect job.
+    Iteratively increment the sample size of the revision being tested, the last
+    known good revision, and the first known bad revision until a relationship
+    of significant difference can be established betweeb the results of the
+    revision being tested and one of the other two.
+
+    If the results do not converge towards finding a significant difference in
+    either direction, this is expected to timeout eventually. This scenario
+    should be rather rare, since it is expected that the fkbr and lkgr are
+    significantly different as a precondition.
 
     Returns:
-      True if this revision is closer to the initial good revision's value than
-      to the initial bad revision's value. False otherwise.
+      True if the results of testing this revision are significantly different
+      from those of testing the earliest known bad revision.
+      False if they are instead significantly different form those of testing
+      the latest knwon good revision.
     """
-    # TODO: Reevaluate this approach
-    bisector = self.bisector
-    distance_to_good = abs(self.mean_value - bisector.good_rev.mean_value)
-    distance_to_bad = abs(self.mean_value - bisector.bad_rev.mean_value)
-    if distance_to_good < distance_to_bad:
-      return True
-    return False
+    diff_from_good = self.bisector.significantly_different(

[dtu, 2016/02/05 00:18:02]

If you reorganize this function, you can avoid the code duplication.

while True:
  diff_from_good = abc
  diff_from_bad = xyz

  if diff_from_good and diff_from_bad:
    # Calculate stuff and return.
  elif diff_from_good or diff_from_bad:
    # Return.
  else:
    # Retest.

+        self.bisector.lkgr.values, self.values)
+    diff_from_bad = self.bisector.significantly_different(
+        self.bisector.fkbr.values, self.values)
+
+    while not (diff_from_good or diff_from_bad):  # pragma: no cover
+      min(self.bisector.lkgr, self, self.bisector.fkbr,
+          key=lambda x: len(x.values)).retest()
+      diff_from_good = self.bisector.significantly_different(
+          self.bisector.lkgr.values, self.values)
+      diff_from_bad = self.bisector.significantly_different(
+          self.bisector.fkbr.values, self.values)
+
+    if diff_from_good and diff_from_bad:
+      # Multiple regressions.
+      # For now, proceed bisecting the biggest difference of the means.
+      dist_from_good = abs(self.mean_value - self.bisector.lkgr.mean_value)
+      dist_from_bad = abs(self.mean_value - self.bisector.fkbr.mean_value)
+      if dist_from_good > dist_from_bad:
+        # TODO(robertocn): Add way to handle the secondary regression
+        #self.bisector.handle_secondary_regression(self, self.bisector.fkbr)
+        return False
+      else:
+        #self.bisector.handle_secondary_regression(self.bisector.lkgr, self)
+        return True
+    return diff_from_bad  # pragma: no cover
 
   def __repr__(self):
     return ('PerfRevisionState(cp=%s, values=%r, mean_value=%r, std_dev=%r)' %
             (self.commit_pos, self.values, self.mean_value, self.std_dev))