Install¶

These instructions target Datoviz v0.4.0rc2, the active release candidate published on PyPI and GitHub. Choose a path by what you want to run.

v0.4 release candidate Python 3.10+ C and C++ Vulkan runtime required

See the v0.4.0rc2 release notes for the package, validation scope, and known limitations. Check project status for the broader release posture.

Choose an installation path¶

You want to...	Use this path now
install the published Python RC	use the exact PyPI command below or a wheel from the release assets
develop a C or C++ application now	Build from source, then follow C/C++ integration
contribute to Datoviz or use the latest `v0.4-dev` code	Build from source
inspect browser support without installing native Datoviz	open a `webgpu-live` route from the Examples section

Native windows and offscreen rendering require a working Vulkan-capable GPU, driver, and runtime. Browser routes use the smaller, experimental WebGPU subset.

RC2 is published

RC2 fixes the packaged macOS native-window Vulkan-loader defect found in RC1. Install RC2 from PyPI for native-window use; RC1 remains available for reproducibility but is superseded.

Python package path¶

Create an isolated environment and install the exact version from PyPI.

macOS / LinuxWindows PowerShell

python -m venv .venv
source .venv/bin/activate
python -m pip install --upgrade pip
python -m pip install --pre datoviz==0.4.0rc2

py -m venv .venv
.\.venv\Scripts\Activate.ps1
py -m pip install --upgrade pip
py -m pip install --pre datoviz==0.4.0rc2

After the final v0.4 package is published, the normal installation command will be:

python -m pip install datoviz

Verify the active environment after installation:

python -c "import datoviz as dvz; print('datoviz import ok')"

Then continue with the Quickstart. Before choosing optional providers or deployment targets, review platform support and known limitations.

C and C++ package path¶

Installed Datoviz packages expose the native library, public headers, runtime assets, and an exported CMake package:

find_package(datoviz CONFIG REQUIRED)
target_link_libraries(my_app PRIVATE datoviz::datoviz)

This CMake fragment is an integration example, not a complete project. The RC2 wheels expose the installed C package and are required to pass the release CMake-consumer checks. Source developers and system packagers may instead follow Build from source, then use the C/C++ integration guide.

Source-build path¶

Build from source when no published package exists, when you need the current development branch, or when you are contributing to Datoviz. The dedicated guide covers prerequisites, shader tools, vendored dependencies, macOS, Linux, native Windows, WSL2, verification, and example commands:

Build Datoviz from source, then return to the Quickstart or First C Program.

Package availability¶

Path	Current v0.4 posture
Python RC package	`datoviz==0.4.0rc2` is published on PyPI and verified against the canonical release wheels.
Source build	Available now for development, C/C++ integration, and package validation.
Native Windows wheels	RC2 AMD64 and ARM64 wheels are published and passed hosted build, package, and clean-install validation.
Source bundle	The RC2 source bundle is published with the GitHub prerelease.
vcpkg and conda-forge	Engineering and preflight paths; not published as RC2 package channels.