/Motif-Discovery

modified exact discovery of time series motifs

Primary LanguageJupyter Notebook

MotifDetection

Exact Discovery of Time Series Motifs

Discovery of motifs in 1-D time series using modified method described by http://www.cs.ucr.edu/~mueen/pdf/EM.pdf

Inputs: Time series: a python list or numpy array, ML: Motif length, K: Major Factor of Clustor Radius, greater than 1 and X, X: Minor Factor of Clustor Radius, greater than 1

Outputs: Motifs: in a list of numpy arrays, BSFB: final euclidean distance calculated

How to use

Copy md.exe and MotifDetection.py to your project folder, and import MotifDetection.py to your main code.

Example

Copy the whole folder and run jupyter notebook file produced as an example

Requirements

  1. Numpy If not installed(windows): pip install numpy

  2. Jupyter Notebook, to run example If not installed(windows): pip install jupyter

  3. Plotly, to run example If not installed(windows): pip install plotly