Skip to content

Implementation of a highly specific test for periodicity of time series based on folding

License

Notifications You must be signed in to change notification settings

neurophysik/periodicitytest

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

12 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Quick and dirty introduction to usage

For an extensive description and explanation, see arXiv:1506.01287.

Very briefly, this test decides whether a given time series can be interpolated by a periodic function with the same number of local extrema as the time series, where only local extrema with a prominence higher than a given error allowance σ are taken into account. The method also returns the period length. It is not intended for experimental but for simulated data.

The method has two parameters that can be chosen straightforwardly:

  • The maximum allowed period length τmax: Choose this parameter reasonably low to limit the runtime and to avoid period lengths that are close to the length of the time series. Unless you know what you are doing, τmax should at least be half the length of the time series.

  • The error allowance σ: Choose this parameter to correspond to the maximum expected absolute numerical error of your simulation, e.g., the maximum absolute integration error of an adaptive integrator.

Instructions for building

The following describes how to use this source depending on how you want to use it:

A. As a Python module

The easiest way is arguably to have the package downloaded and installed by Pip, for example:

pip3 install periodicitytest

Otherwise, build and install the module by running (for example)

python3 setup.py install --user

In both cases, you should end up with a Python module called periodicitytest, which contains one function called periodicitytest. It can be loaded, e.g., as follows:

from periodicitytest import periodicitytest

For further documentation, see periodicitytest's docstring.

B. As a standalone program

Build the program by running

python3 setup_C.py

or

python3 setup_C.py test

if you want to run the tests.

This generates an executable called periodicitytest in the folder bin. It takes the maximum period length τmax and the error allowance σ as an argument. The time series is read from STDIN.

Take a look at standalone.c if you want to modify input and output.

C. As a C library

The central function is contained and documented in search.h. For input and output it requires simple datatypes defined in basics_standalone.h and interval.h. It needs to compiled with -DSTANDALONE. Several costly assertions can be avoided by compiling with -DNDEBUG.

About

Implementation of a highly specific test for periodicity of time series based on folding

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published