N-Dim-view plugin for dpVision

N-Dim-view is a plugin for the open-source platform dpVision that provides an interactive 3D visualisation of multidimensional data using expressive avatars. The plugin combines spatial projection with glyph-based visual metaphors to help users explore complex datasets in an intuitive and visually engaging way.

🧠 Key Features

• Chernoff-inspired avatars – Each data point is visualised as a small figure ("avatar") whose facial traits represent selected data dimensions.
• 3D spatial projection – Additional dimensions are mapped to position in a 3D coordinate system, allowing for interactive navigation and inspection.
• Custom slab slicing – Users can control the thickness of the visible data layer in nD space, focusing on cross-sections of interest.
• Manual or automatic assignment – Flexible mapping of data dimensions to avatar appearance or spatial coordinates.
• Support for real and synthetic data – Tested with a 15D wine quality dataset and an artificial 4D tesseract-shaped dataset.

🧑‍💻 Author and Contributors

The main author of this plugin is Leszek Luchowski, responsible for the concept, model design, and algorithmic implementation.

Dariusz Pojda contributed to the project by integrating the plugin with the dpVision platform and optimising its performance.

📦 Installation

The plugin is designed to work with dpVision version X.X or higher.

1. Clone this repository:

   git clone https://github.com/iitis/n-dim-view.git

2. Copy the plugin folder to the dpVision/plugins/ directory.

3. Reconfigure your dpVision project with CMake or — if you do not use CMake — simply add a plugin project (n-dim-view.vcxproj) to dpVision solution in MS Visual Studio

4. Build your dpVision project

5. Launch dpVision and find the plugin in the plugin manager.

6. Use the plugin with your dataset.

🚀 Example Use Cases

• Visualising clusters in high-dimensional data
• Exploring correlations between human-perceived qualities and technical parameters
• Educational visualisation of abstract geometries (e.g., 4D rotation)
• Dimensionality reduction with interpretability

🛠️ Roadmap

Planned features include:

• Support for alternative glyph styles (trees, neutral geometric forms)
• Heuristic or ML-based automatic dimension-to-feature mapping
• Scalability enhancements for large datasets
• Export to interactive WebGL viewers
• User feedback and usability testing

📄 License

This plugin is released under the LGPL-2.1 License. See LICENSE for details.