Signal Processing Algorithms from MongoDB
Signal Processing Algorithms
Getting Started - Users
pip install signal-processing-algorithms
Getting Started - Developers
Getting the code:
$ git clone firstname.lastname@example.org:mongodb/signal-processing-algorithms.git $ cd signal-processing-algorithms
Making a virtual environment and installing the stuff you need into it:
$ virtualenv -p python3 venv $ source venv/bin/activate $ pip install -e . $ pip install -r requirements.txt
Running the slow tests:
$ pytest --runslow
Some of the larger tests can take a significant amount of time (more than 2 hours).
$ pytest src --flake8 --black --pydocstyle --mypy
Intro to E-Divisive
Detecting distributional changes in a series of numerical values can be surprisingly difficult. Simple systems based on thresholds or mean values can be yield false positives due to outliers in the data, and will fail to detect changes in the noise profile of the series you are analyzing.
One robust way of detecting many of the changes missed by other approaches is to use E-Divisive with Means, an energy statistic based approach that compares the expected distance (Euclidean norm) between samples of two portions of the series with the expected distance between samples within those portions.
That is to say, assuming that the two portions can each be modeled as i.i.d. samples drawn from distinct random variables (X for the first portion, Y for the second portion), you would expect the following to be non-zero if there is a sdifference between the two portions:
Where alpha is some fixed constant in (0, 2). This can be calculated empirically with samples from the portions corresponding to X, Y as follows:
Thus for a series Z of length L, we find the most likely change point by solving the following for argmax(τ) (with a scaling factor of mn/(m+n) and α=1 for simplicity):
Multiple Change Points
The algorithm for finding multiple change points is also simple.
Assuming you have some k known change points:
- Partition the series into segments between/around these change points.
- Find the maximum value of our divergence metric within each partition.
- Take the maximum of the maxima we have just found --> this is our k+1th change point.
- Return to step 1 and continue until reaching your stopping criterion.
In this package we have implemented a permutation based test as a stopping criterion:
After step 3 of the multiple change point procedure above, randomly permute all of the data within each cluster, and find the most likely change point for this permuted data using the procedure laid out above.
After performing this operation z times, count the number of permuted change points z' that have higher divergence metrics than the change point you calculated with un-permuted data. The significance level of your change point is thus z'/(z+1).
We allow users to configure a permutation tester with
permutations representing the significance cutoff for algorithm termination and permutations to perform for each
from signal_processing_algorithms.e_divisive import EDivisive from signal_processing_algorithms.e_divisive.calculators import cext_calculator from signal_processing_algorithms.e_divisive.significance_test import QHatPermutationsSignificanceTester from some_module import series // Use C-Extension calculator for calculating divergence metrics calculator = cext_calculator // Permutation tester with 1% significance threshold performing 100 permutations for each change point candidate tester = QHatPermutationsSignificanceTester( calculator=calculator, pvalue=0.01, permutations=100 ) algo = EDivisive(calculator=calculator, significance_tester=tester) change_points = algo.get_change_points(series)
Release history Release notifications
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
|Filename, size||File type||Python version||Upload date||Hashes|
|Filename, size signal_processing_algorithms-1.0.4-cp37-cp37m-macosx_10_14_x86_64.whl (21.4 kB)||File type Wheel||Python version cp37||Upload date||Hashes View|
|Filename, size signal_processing_algorithms-1.0.4.tar.gz (14.7 kB)||File type Source||Python version None||Upload date||Hashes View|
Hashes for signal_processing_algorithms-1.0.4-cp37-cp37m-macosx_10_14_x86_64.whl
Hashes for signal_processing_algorithms-1.0.4.tar.gz