geoist.snoopy package

Subpackages

• geoist.snoopy.algorithms package
  • Subpackages
    • geoist.snoopy.algorithms.anomaly_detector_algorithms package
      • Submodules
      • geoist.snoopy.algorithms.anomaly_detector_algorithms.absolute_threshold module
      • geoist.snoopy.algorithms.anomaly_detector_algorithms.all module
      • geoist.snoopy.algorithms.anomaly_detector_algorithms.bitmap_detector module
      • geoist.snoopy.algorithms.anomaly_detector_algorithms.default_detector module
      • geoist.snoopy.algorithms.anomaly_detector_algorithms.derivative_detector module
      • geoist.snoopy.algorithms.anomaly_detector_algorithms.diff_percent_threshold module
      • geoist.snoopy.algorithms.anomaly_detector_algorithms.exp_avg_detector module
      • geoist.snoopy.algorithms.anomaly_detector_algorithms.sign_test module
      • Module contents
    • geoist.snoopy.algorithms.correlator_algorithms package
      • Submodules
      • geoist.snoopy.algorithms.correlator_algorithms.all module
      • geoist.snoopy.algorithms.correlator_algorithms.cross_correlator module
      • Module contents
  • Module contents
• geoist.snoopy.modules package
  • Submodules
  • geoist.snoopy.modules.anomaly module
  • geoist.snoopy.modules.correlation_result module
  • geoist.snoopy.modules.time_series module
  • Module contents

Submodules

geoist.snoopy.anomaly_detector module

API for Anomaly Detector Module This module detects anomalies in a single time series.

class geoist.snoopy.anomaly_detector.AnomalyDetector(time_series, baseline_time_series=None, score_only=False, score_threshold=None, score_percent_threshold=None, algorithm_name=None, algorithm_params=None, refine_algorithm_name=None, refine_algorithm_params=None, algorithm_class=None)

基类：object

get_all_scores()

Get anomaly scores. :return: a TimeSeries object represents anomaly scores.

get_anomalies()

Get anomalies. :return list: a list of Anomaly objects.

geoist.snoopy.constants module

Constants to use for geoist.snoopy

geoist.snoopy.constants.DEFAULT_DETECTOR_EMA_SIGNIFICANT = 0.94

Correlator Constants

geoist.snoopy.correlator module

API for Correlator Module This module finds correlation between two time series.

class geoist.snoopy.correlator.Correlator(time_series_a, time_series_b, time_period=None, use_anomaly_score=False, algorithm_name=None, algorithm_params=None)

基类：object

get_correlation_result()

Get correlation result. :return CorrelationResult: a CorrelationResult object.

is_correlated(threshold=0)

Compare with a threshold to determine whether two timeseries correlate to each other. :return: a CorrelationResult object if two time series correlate otherwise false.

geoist.snoopy.exceptions module

Exception Classes

exception geoist.snoopy.exceptions.AlgorithmNotFound

基类：Exception

Raise when algorithm can not be found.

exception geoist.snoopy.exceptions.InvalidDataFormat

基类：Exception

Raise when data has invalid format.

exception geoist.snoopy.exceptions.NotEnoughDataPoints

基类：Exception

Raise when there are not enough data points.

exception geoist.snoopy.exceptions.RequiredParametersNotPassed

基类：Exception

Raise when algorithm can not be properly initialized because some required parameters are not passed in init.

geoist.snoopy.tsa module

geoist.snoopy.tsa.dejump(dt, th=None, fill_end=True, *args, **kwargs)

geoist.snoopy.tsa.departure(data, freq='M')

Calculate departure(ju4ping2 in Chinese)

参数

• data (Series or 1-column DataFrame) -- Data to be processed

• freq (str) -- M==Month,D==Day

返回

DataFrame

geoist.snoopy.tsa.despike(dt, th=None, fill_end=True, *args, **kwargs)

geoist.snoopy.tsa.despike_v2(dt, th=None, fill_end=True, *args, **kwargs)

geoist.snoopy.tsa.print_adf(res, data_name, file=<colorama.ansitowin32.StreamWrapper object>)

geoist.snoopy.utils module

Utilities for snoopy

geoist.snoopy.utils.compute_ema(smoothing_factor, points)

Compute exponential moving average of a list of points. :param float smoothing_factor: the smoothing factor. :param list points: the data points. :return list: all ema in a list.

geoist.snoopy.utils.pbinom(k, n)

Compute cdf for binomial with prob = 0.5 compare to R pbinom :param k: :param n: :return: cumulative probability

geoist.snoopy.utils.qbinom(p, n)

quantile function for binomial with probability of success 0.5 returns smallest k such that Prob(X <= k) >= p compare to R qbinom :param n: number :param p: quantile level :return: k

geoist.snoopy.utils.read_csv(csv_name)

Read data from a csv file into a dictionary. :param str csv_name: path to a csv file. :return dict: a dictionary represents the data in file.

geoist.snoopy.utils.to_epoch(t_str)

Covert a timestamp string to an epoch number. :param str t_str: a timestamp string. :return int: epoch number of the timestamp.

Module contents

class geoist.snoopy.Snoopy(anomalies, correlations)

基类：object

get_root_causes()

Get root causes. :return dict: a dict represents root causes for each anomaly.