SAX Demo

This Dash application allows experimenting with the parameters of the SAX transformation. For more information on the transformation, please refer to Sec. 2.3 of iSAX: Indexing and Mining Terabyte Sized Time Series, Jin Shieh & Eamonn Keogh, SIGKDD, 2008

By default, a timeseries of length 128 is used. For comparison, the code also contains a shorter time series which is very similar to the one in the above-mentioned paper (see line sax_dash.py:130).

Instructions Python 3.9.2

Use requirements.txt to install dependencies. Tested with Python 3.9.2 / Windows.

Instructions Anaconda

After cloning the repository, run the following commands

1. conda create -n saxts dash numpy scipy
2. conda activate saxts
3. python sax_dash.py
4. Use your browser to navigate to http://127.0.0.1:8050/