PolarViz, a radial distortion strategy based on RadViz plot[1], proposed by the authors to visualize high-dimensional dataset [2][3]. The user can customize the radial distribution of data points in RadViz plot by operations on a related histogram plot. This program is in JavaScript and only depend on d3.js library. The implementation of RadViz in JavaScript and d3.js can be found here (one by me and one by Chris Viau)
To be honest, this is my first github project and even the first JavaScript project. Though the code works well in the demo, I cannot guarantee the performance on any scenarios. I will keep modifying this code.
If this work makes any contribution to your own work, I would be very happy. If you think your work has any relationship with PolarViz and would like to cite this work, thanks. For your convenience, I provide the Latex BibTex of this work.
@inproceedings{wang2017histogram,
title={Histogram equalization and specification for high-dimensional data visualization using RadViz},
author={Wang, Yan-Chao and Zhang, Qian and Lin, Feng and Goh, Chi-Keong and Wang, Xuan and Seah, Hock-Soon},
booktitle={Proceedings of the Computer Graphics International Conference},
pages={15},
year={2017},
organization={ACM}
}
@article{wang2018polarviz,
title={PolarViz: a discriminating visualization and visual analytics tool for high-dimensional data},
author={Wang, Yan Chao and Zhang, Qian and Lin, Feng and Goh, Chi Keong and Seah, Hock Soon},
journal={The Visual Computer},
pages={1--16},
year={2018},
publisher={Springer}
}
- 'Index.html' defines the DOM components used in the PolarViz function.
- 'Index.js' loads the high-dimensional dataset, defines the five variables used in the PolarViz function, and calls the PolarViz function.
- 'PolarViz.js' defines the PolarViz function.
- In the data file, I use the 'IRIS' dataset as an example to call the PolarViz function.