MulTiDR: A Visual Analytics Framework for Reviewing Multivariate Time-Series Data with Dimensionality Reduction

New

• Now, MulTiDR is available from all major OSs, Mac OS, Linux, and Windows (but the websocket server implementation used for the web UI is beta for Windows)

About

• MulTiDR is from: Fujiwara et al., "A Visual Analytics Framework for ReviewingMultivariate Time-Series Data with Dimensionality Reduction." IEEE Transactions on Visualization and Computer Graphics, vol. 27, no. 2, pp. 1601-1611, 2021.

• Implementation of MulTiDR back-end algorithms

• Two-step DR (TDR): Framework of dimensionality reduction for multivariate time-series data. TDR produces a low-dimensional representation from a third-order tensor.

• Contrastive learning with sign adjustment of feature contributions.

• Implementation of MulTiDR Visual Interface.

• Demonstration video of a system using MulTiDR: https://takanori-fujiwara.github.io/s/multidr/

---

Back-End Library Setup

Requirements

• Python3
• Note: Tested on macOS Sequoia, Ubuntu 22.0.4 LTS, and Windows 10.

Setup

• Install with pip3. Move to the directory of this repository. Then,

  pip3 install .

• If you want to use contrastive learning with a default setting (i.e., use of ccPCA), install ccPCA from: https://github.com/takanori-fujiwara/ccpca

Usage

• Import installed modules from python (e.g., from multidr.tdr import TDR). See sample.py for examples.
• For detailed documentations, please see doc/index.html or directly see comments in multidr/tdr.py and multidr/cl.py.

---

Web-based Visual Interface Setup

Requirements

• Server side

  • Python3
  • HTTP Server

• Client side

  • Browser supporting JavaScript ES2015(ES6) and WebGL 2.0.
  • Internet connection (to access D3 library)

• Note: Tested on macOS Ventura, Ubuntu 22.0.4 LTS, and Windows 10.

Server Setup

• Install multidr and ccpca modules based on "Back-End Library Setup"

• Move to ui/server/ of this repository. Then,

  pip3 install -r requirements.txt

  (This intalls numpy, scipy, uvloop, websockets)

• Run websocket server:

  python3 ws_server.py or python ws_server.py

• Run http server. For example, move to ui/client/ of this repository. Then,

  python3 -m http.server or python -m http.server

Client Setup

• Access to the url setup in the http server. For example, if you set an http server with the above command. You can acess with: http://localhost:8000/

How to Include New Datasets

• Please, refer to ui/doc/data_format.md

---

How to Cite

Please, cite:
Takanori Fujiwara, Shilpika, Naohisa Sakamoto, Jorji Nonaka, Keiji Yamamoto, and Kwan-Liu Ma, "A Visual Analytics Framework for ReviewingMultivariate Time-Series Data with Dimensionality Reduction". IEEE Transactions on Visualization and Computer Graphics, vol. 27, no. 2, pp. 1601-1611, 2021.