iPyParaView

Learn more at: https://developer.nvidia.com/gtc/2020/video/s22111

A widget for interactive server-side ParaView rendering.

Examples

Example notebooks are avaible in the notebooks folder. They are designed to give a broad overview of how to use ipyparaview. New users will probably have the best luck jumping in to the Hello_Jupyter-ParaView.ipynb notebook, which demonstrates basic usage and setting up the ParaView display. The Iso-Surfaces_with_RTX.ipynb notebook demonstrates more advanced usage, with more extensive manipulation of the render state and interactive control. The Dask-MPI_Volume_Render.ipynb notebook demonstrates how to use multi-node rendering by running PVRenderActors on a Dask-MPI cluster.

Installation

There are two ways to install ipyparaview, both of which rely on pip. The lightweight method is to point pip at the GitHub repo, and then enable the notebook extension (note: this will not enable the extension for Jupyter lab). The more fully-featured method is to download a copy of the source code and install from the local version. We typically run inside of a conda environment (see conda setup step in instructions for the full-feature install with conda)

Lightweight quick install

$ pip install git+https://github.com/Kitware/ipyparaview.git
$ jupyter nbextension enable --py --sys-prefix ipyparaview

From within a conda environment:

$ conda env create -f environment.yml
$ conda activate ipy_pv_dev
$ ./rebuild.sh

Fully-featured local source installation

$ git clone https://github.com/NVIDIA/ipyparaview.git
$ cd ipyparaview
$ ./build.sh

Fully-featured local source installation (with conda)

$ git clone https://github.com/Kitware/ipyparaview.git
$ cd ipyparaview
$ pip install -e .
$ jupyter nbextension install --py --symlink --sys-prefix ipyparaview
$ jupyter nbextension enable --py --sys-prefix ipyparaview
$ jupyter labextension install js

Running

Within a conda environment

$ conda activate ipy_pv_dev
$ export LD_LIBRARY_PATH=$PVPATH/lib/
$ export PYTHONPATH=$PVPATH/lib/python3.7/site-packages/
$ jupyter notebook

Or from a Docker container, create an image by:

$ docker build -t ipp_base -f base.dockerfile .
$ docker build -t ipp .

Then run that container by:

$ docker run -p 8888:8888 ipp

Demos

Our conda environment installs all required dependencies for our demos.

conda activate ipy_pv_dev
export PYTHONPATH=$PVPATH/lib/python3.7/site-packages/ # or `conda install paraview`
cd notebooks/
jupyter notebook