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
Copy md.exe and MotifDetection.py to your project folder, and import MotifDetection.py to your main code.
Copy the whole folder and run jupyter notebook file produced as an example
-
Numpy If not installed(windows): pip install numpy
-
Jupyter Notebook, to run example If not installed(windows): pip install jupyter
-
Plotly, to run example If not installed(windows): pip install plotly