aboutsummaryrefslogtreecommitdiffstats
path: root/resources/tools/presentation/utils.py
diff options
context:
space:
mode:
authorVratko Polak <vrpolak@cisco.com>2018-06-20 12:56:41 +0200
committerVratko Polak <vrpolak@cisco.com>2018-06-20 12:56:41 +0200
commit2e63ef13b419da1198439617e66cb0f1cfe6be65 (patch)
tree634ba3c2134a3ed8c81f8ebdc1a9c4a2ef14712b /resources/tools/presentation/utils.py
parent39b4a07718ecab94ea331362edb62dfcf678bd09 (diff)
CSIT-1110: Replace old trending with the new one
+ Remove /new/ folders in presentation and docs. Change-Id: I870002ba8509189196e778aa1292b93e83a3ec17 Signed-off-by: Vratko Polak <vrpolak@cisco.com>
Diffstat (limited to 'resources/tools/presentation/utils.py')
-rw-r--r--resources/tools/presentation/utils.py144
1 files changed, 40 insertions, 104 deletions
diff --git a/resources/tools/presentation/utils.py b/resources/tools/presentation/utils.py
index ba32932187..a2aa0dc071 100644
--- a/resources/tools/presentation/utils.py
+++ b/resources/tools/presentation/utils.py
@@ -17,7 +17,6 @@
import multiprocessing
import subprocess
import numpy as np
-import pandas as pd
import logging
import csv
import prettytable
@@ -28,6 +27,7 @@ from shutil import move, Error
from math import sqrt
from errors import PresentationError
+from jumpavg.BitCountingClassifier import BitCountingClassifier
def mean(items):
@@ -71,73 +71,6 @@ def relative_change(nr1, nr2):
return float(((nr2 - nr1) / nr1) * 100)
-def remove_outliers(input_list, outlier_const=1.5, window=14):
- """Return list with outliers removed, using split_outliers.
-
- :param input_list: Data from which the outliers will be removed.
- :param outlier_const: Outlier constant.
- :param window: How many preceding values to take into account.
- :type input_list: list of floats
- :type outlier_const: float
- :type window: int
- :returns: The input list without outliers.
- :rtype: list of floats
- """
-
- data = np.array(input_list)
- upper_quartile = np.percentile(data, 75)
- lower_quartile = np.percentile(data, 25)
- iqr = (upper_quartile - lower_quartile) * outlier_const
- quartile_set = (lower_quartile - iqr, upper_quartile + iqr)
- result_lst = list()
- for y in input_list:
- if quartile_set[0] <= y <= quartile_set[1]:
- result_lst.append(y)
- return result_lst
-
-
-def split_outliers(input_series, outlier_const=1.5, window=14):
- """Go through the input data and generate two pandas series:
- - input data with outliers replaced by NAN
- - outliers.
- The function uses IQR to detect outliers.
-
- :param input_series: Data to be examined for outliers.
- :param outlier_const: Outlier constant.
- :param window: How many preceding values to take into account.
- :type input_series: pandas.Series
- :type outlier_const: float
- :type window: int
- :returns: Input data with NAN outliers and Outliers.
- :rtype: (pandas.Series, pandas.Series)
- """
-
- list_data = list(input_series.items())
- head_size = min(window, len(list_data))
- head_list = list_data[:head_size]
- trimmed_data = pd.Series()
- outliers = pd.Series()
- for item_x, item_y in head_list:
- item_pd = pd.Series([item_y, ], index=[item_x, ])
- trimmed_data = trimmed_data.append(item_pd)
- for index, (item_x, item_y) in list(enumerate(list_data))[head_size:]:
- y_rolling_list = [y for (x, y) in list_data[index - head_size:index]]
- y_rolling_array = np.array(y_rolling_list)
- q1 = np.percentile(y_rolling_array, 25)
- q3 = np.percentile(y_rolling_array, 75)
- iqr = (q3 - q1) * outlier_const
- low = q1 - iqr
- item_pd = pd.Series([item_y, ], index=[item_x, ])
- if low <= item_y:
- trimmed_data = trimmed_data.append(item_pd)
- else:
- outliers = outliers.append(item_pd)
- nan_pd = pd.Series([np.nan, ], index=[item_x, ])
- trimmed_data = trimmed_data.append(nan_pd)
-
- return trimmed_data, outliers
-
-
def get_files(path, extension=None, full_path=True):
"""Generates the list of files to process.
@@ -276,46 +209,49 @@ def archive_input_data(spec):
logging.info(" Done.")
-def classify_anomalies(data, window):
- """Evaluates if the sample value is an outlier, regression, normal or
- progression compared to the previous data within the window.
- We use the intervals defined as:
- - regress: less than trimmed moving median - 3 * stdev
- - normal: between trimmed moving median - 3 * stdev and median + 3 * stdev
- - progress: more than trimmed moving median + 3 * stdev
- where stdev is trimmed moving standard deviation.
+def classify_anomalies(data):
+ """Process the data and return anomalies and trending values.
+
+ Gather data into groups with average as trend value.
+ Decorate values within groups to be normal,
+ the first value of changed average as a regression, or a progression.
- :param data: Full data set with the outliers replaced by nan.
- :param window: Window size used to calculate moving average and moving
- stdev.
+ :param data: Full data set with unavailable samples replaced by nan.
:type data: pandas.Series
- :type window: int
- :returns: Evaluated results.
- :rtype: list
+ :returns: Classification and trend values
+ :rtype: 2-tuple, list of strings and list of floats
"""
-
- if data.size < 3:
- return None
-
- win_size = data.size if data.size < window else window
- tmm = data.rolling(window=win_size, min_periods=2).median()
- tmstd = data.rolling(window=win_size, min_periods=2).std()
-
- classification = ["normal", ]
- first = True
- for build, value in data.iteritems():
- if first:
- first = False
- continue
- if np.isnan(value) or np.isnan(tmm[build]) or np.isnan(tmstd[build]):
+ # Nan mean something went wrong.
+ # Use 0.0 to cause that being reported as a severe regression.
+ bare_data = [0.0 if np.isnan(sample) else sample
+ for _, sample in data.iteritems()]
+ # TODO: Put analogous iterator into jumpavg library.
+ groups = BitCountingClassifier().classify(bare_data)
+ groups.reverse() # Just to use .pop() for FIFO.
+ classification = []
+ avgs = []
+ active_group = None
+ values_left = 0
+ avg = 0.0
+ for _, sample in data.iteritems():
+ if np.isnan(sample):
classification.append("outlier")
- elif value < (tmm[build] - 3 * tmstd[build]):
- classification.append("regression")
- elif value > (tmm[build] + 3 * tmstd[build]):
- classification.append("progression")
- else:
- classification.append("normal")
- return classification
+ avgs.append(sample)
+ continue
+ if values_left < 1 or active_group is None:
+ values_left = 0
+ while values_left < 1: # Ignore empty groups (should not happen).
+ active_group = groups.pop()
+ values_left = len(active_group.values)
+ avg = active_group.metadata.avg
+ classification.append(active_group.metadata.classification)
+ avgs.append(avg)
+ values_left -= 1
+ continue
+ classification.append("normal")
+ avgs.append(avg)
+ values_left -= 1
+ return classification, avgs
def convert_csv_to_pretty_txt(csv_file, txt_file):