CarnaPy

The aim of this package is to provide real-time 3D visualization in Python for specifically, but not limited to, biomedical data. The library is based on Carna.

See examples/kalinin2018.ipynb for an example.

---

Contents

• Limitations
• Dependencies
• Installation
• Build instructions

---

1. Limitations

• Only 8bit and 16bit volume data are supported at the moment.
• DRR renderings are not exposed to Python yet.
• Build process is currently limited to Linux-based systems.

---

2. Dependencies

Using the library requires the following dependencies:

• numpy ≥ 1.16
• EGL driver support
• OpenGL 3.3
• Python ≥ 3.7

The following dependencies must be satisfied for the build process:

• Carna ≥ 3.1
• Eigen ≥ 3.0.5
• libboost-iostreams
• pybind11
• EGL development files

In addition, the following dependencies are required to run the test suite:

• matplotlib
• scipy

---

3. Installation

The easiest way to install and use the library is to use one of the binary Conda packages:

conda install -c kostrykin carnapy

Conda packages are available for Python 3.7–3.9.

---

4. Build instructions

Assuming you are using a recent version of Ubuntu:

sudo apt-get -qq install libegl1-mesa-dev libboost-iostreams-dev

Create and activate a Conda environment to work in, then:

conda install -c conda-forge pybind11

Grab a recent version of Eigen, unpack it, and tell CMake where it is located:

wget https://gitlab.com/libeigen/eigen/-/archive/3.2.10/eigen-3.2.10.tar.gz
tar -vzxf eigen-3.2.10.tar.gz -C /tmp/
export CMAKE_PREFIX_PATH="/tmp/eigen-3.2.10:$CMAKE_PREFIX_PATH"

If you have not already, download, build, and install Carna:

git clone git@github.com:kostrykin/Carna.git build_carna
cd build_carna
sh linux_build.sh

Now it is time to build, package, and install CarnaPy:

cd ..
python setup.py build
python setup.py install