Lint bisect-related modules; obey pylint.

scripts/slave/recipe_modules/math_utils/api.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.
 
 """General statistical or mathematical functions."""
 
 import math
 
 from slave import recipe_api
 
+
 class MathUtilsApi(recipe_api.RecipeApi):
 
   @staticmethod
   def truncated_mean(data_set, truncate_fraction):
     """Calculates the truncated mean of a set of values.
 
     Note that this isn't just the mean of the set of values with the highest
     and lowest values discarded; the non-discarded values are also weighted
     differently depending how many values are discarded.
 
@@ skipping 110 matching lines @@
   def standard_error(values):
     """Calculates the standard error of a list of values."""
     if len(values) <= 1:
       return 0.0
     std_dev = MathUtilsApi.standard_deviation(values)
     return std_dev / math.sqrt(len(values))
 
   #Copied this from BisectResults
   @staticmethod
   def confidence_score(sample1, sample2,
-                      accept_single_bad_or_good=False):
+                       accept_single_bad_or_good=False):
     """Calculates a confidence score.
 
     This score is a percentage which represents our degree of confidence in the
     proposition that the good results and bad results are distinct groups, and
     their differences aren't due to chance alone.
 
 
     Args:
       sample1: A flat list of "good" result numbers.
       sample2: A flat list of "bad" result numbers.
@@ skipping 92 matching lines @@
     if v1 == 0 and v2 == 0:
       return 1
     # If the sample size is too small, also return the minimum (1).
     if n1 <= 1 or n2 <= 2:
       return 1
     df = (((v1 / n1 + v2 / n2) ** 2) /
           ((v1 ** 2) / ((n1 ** 2) * (n1 - 1)) +
            (v2 ** 2) / ((n2 ** 2) * (n2 - 1))))
     return max(1, df)
 
-
   # Below is a hard-coded table for looking up p-values.
   #
   # Normally, p-values are calculated based on the t-distribution formula.
   # Looking up pre-calculated values is a less accurate but less complicated
   # alternative.
   #
   # Reference: http://www.sjsu.edu/faculty/gerstman/StatPrimer/t-table.pdf
 
   # A list of p-values for a two-tailed test. The entries correspond to to
   # entries in the rows of the table below.
@@ skipping 83 matching lines @@
       lesser_equal = [n for n in nums if n <= target]
       assert lesser_equal, 'No number in number list <= target.'
       return max(lesser_equal)
 
     df_key = greatest_smaller(MathUtilsApi.TABLE.keys(), df)
     t_table_row = MathUtilsApi.TABLE[df_key]
     approximate_t_value = greatest_smaller(t_table_row, t)
     t_value_index = t_table_row.index(approximate_t_value)
 
     return MathUtilsApi.TWO_TAIL[t_value_index]