d_{symb} playground

A fast interactive exploration of multivariate time series datasets

Try our demo

$d_{symb}$ playground is a Python-based web interactive tool to interpet and compare large multivariate time series datasets. It is based on a novel symbolic representation, called $d_{symb}$, for multivariate time series. $d_{symb}$ allows to visualize a dataset of multivariate time series with a single glance, thus to quickly gain insights on your data. $d_{symb}$ also comes with a compatible distance measure to compare the obtained symbolic sequences. Apart from its relevance on data mining tasks, this distance measure is also fast. Indeed, comparing a dataset of 80 time series (with 80 dimensions and 5,000 timestamps) requires 20 seconds instead of 2,000 seconds for DTW-based analysis.

Reference

This repository contains the code that supports the following publication on the $d_{symb}$ playground.

Demo paper of the $d_{symb}$ playground [paper / PDF / Streamlit app / 4 min YouTube video]:

S. W. Combettes, P. Boniol, C. Truong, and L. Oudre. d_{symb} playground: an interactive tool to explore large multivariate time series datasets. In Proceedings of the International Conference on Data Engineering (ICDE) (to appear), Utrecht, Netherlands, 2024.

@inproceedings{2024_combettes_dsymb_playground_icde,
  title={d_{symb} playground: an interactive tool to explore large multivariate time series datasets},
  author={Sylvain W. Combettes and Paul Boniol and Charles Truong and Laurent Oudre},
  booktitle={Proceedings of the International Conference on Data Engineering (ICDE) (to appear)},
  year={2024},
  location={Utrecht, Netherlands},
}

Method paper of $d_{symb}$ [paper / PDF / code]:

S. W. Combettes, C. Truong, and L. Oudre. An Interpretable Distance Measure for Multivariate Non-Stationary Physiological Signals. In Proceedings of the International Conference on Data Mining Workshops (ICDMW), Shanghai, China, 2023.

@inproceedings{2023_combettes_dsymb_icdm,
  author={Combettes, Sylvain W. and Truong, Charles and Oudre, Laurent},
  booktitle={2023 IEEE International Conference on Data Mining Workshops (ICDMW)}, 
  title={An Interpretable Distance Measure for Multivariate Non-Stationary Physiological Signals}, 
  year={2023},
  pages={533-539},
  doi={10.1109/ICDMW60847.2023.00076},
  location={Shanghai, China},
}

Contributors

• Sylvain W. Combettes (Centre Borelli, ENS Paris-Saclay)
• Paul Boniol (Centre Borelli, ENS Paris-Saclay)
• Charles Truong (Centre Borelli, ENS Paris-Saclay)
• Laurent Oudre (Centre Borelli, ENS Paris-Saclay)

Usage

Step 1: Clone this repository using git and change into its root directory.

git clone https://github.com/boniolp/dsymb-playground.git
cd dsymb-playground/

Step 2: Create and activate a conda environment and install the dependencies.

conda create -n dsymb-playground python=3.9
conda activate dsymb-playground
pip install -r requirements.txt

Step 3: You can use our tool in two different ways:

• Access online: https://dsymb-playground.streamlit.app/
• Run locally (preferable for large time series datasets). To do so, run the following command:

streamlit run app.py

You can then open the app using your web browser. You can upload any kind of time series (one file per time series) with the shape (n_timestamps, n_dims). A preprocessed version of the dataset JIGSAWS dataset can be found here.

Acknowledgments

Sylvain W. Combettes is supported by the IDAML chair (ENS Paris-Saclay) and UDOPIA (ANR-20-THIA-0013-01). Charles Truong is funded by the PhLAMES chair (ENS Paris-Saclay). Part of the computations has been executed on Atos Edge computer, funded by the IDAML chair (ENS Paris-Saclay).