Iteratively increase sample size for good/bad classification.

scripts/slave/recipe_modules/auto_bisect/bisector.py

Index: scripts/slave/recipe_modules/auto_bisect/bisector.py
diff --git a/scripts/slave/recipe_modules/auto_bisect/bisector.py b/scripts/slave/recipe_modules/auto_bisect/bisector.py
index 809d33bdaeba8457293d0a36e132c12dd2abd434..4220a66c04da20d27f45439da771bffe9cd0fd3e 100644
--- a/scripts/slave/recipe_modules/auto_bisect/bisector.py
+++ b/scripts/slave/recipe_modules/auto_bisect/bisector.py
@@ -114,6 +114,22 @@ class Bisector(object):
     if init_revisions:
       self._expand_chromium_revision_range()
 
+  def significantly_different(self, list_a, list_b,
+                              significance_level=0.05): # pragma: no cover
+    step_result = self.api.m.python(
+        'Checking sample difference',
+        self.api.resource('significantly_different.py'),
+        [json.dumps(list_a), json.dumps(list_b), str(significance_level)],
+        stdout=self.api.m.json.output())
+    results = step_result.stdout
+    if results is None:
+      assert self.dummy_builds
+      return True
+    significantly_different = results['significantly_different']
+    step_result.presentation.logs[str(significantly_different)] = [
+        'See json.output for details']

[qyearsley, 2016/01/26 19:26:00]

How will this show up on the build page? Will there be a log link (along with
stdio and json.output) named either "True" or "False"?

[RobertoCN, 2016/02/01 17:29:50]

https://goo.gl/photos/E9AuQkm9C8Mpjyvw7

+    return significantly_different
+
   def config_step(self):
     """Yields a simple echo step that outputs the bisect config."""
     api = self.api