Datoviz is a high-performance GPU-powered visualization engine designed for interactive scientific data. It addresses the need for greater control and performance when visualizing dense points, images, meshes, volumes, and complex scientific scenes, surpassing the capabilities of ordinary plotting libraries. Primarily targeting researchers and power users in scientific domains, Datoviz offers a robust solution for demanding visualization tasks.

How It Works

Datoviz employs a retained scene model and native Vulkan rendering for desktop and offscreen presentation. This approach allows for efficient handling of large datasets and complex scenes by leveraging the GPU. Python users can directly integrate NumPy arrays into the visualization pipeline, while C/C++ applications can utilize the native C API. The engine focuses on reproducible capture and provides direct use from both Python and C.

Quick Start & Requirements

Installation for the release candidate is python -m pip install --pre datoviz==0.4.0rc2. After the final v0.4 release, use python -m pip install datoviz. Building from source requires Git, CMake 3.21+, Ninja, just, Python 3.10+, a C/C++ compiler, shader tools, and a Vulkan-capable runtime. The full installation guide and an example gallery are mentioned but not directly linked.

Highlighted Details

• v0.4 introduces a retained scene model, native Vulkan rendering, and supports desktop/offscreen presentation.
• Direct integration with Python NumPy arrays and a native C/C++ API are supported.