Download OSCARS 1.36.12

Installation - Basic

pip install oscars -U

Simple Installation from Source

The simplest way to install OSCARS from source is using conda. This guarantees that you have all of the correct compilers and libraries. If you don't want to bother with conda, you can try the non-conda directions that follow

Installing with conda

# Install git in your root conda environment (if you don't already have it).
conda install -n root git

# Fix for temporary conda bug for some distributions
conda install -n root pyyaml

# Download the materials.
git clone https://github.com/dhidas/OSCARS -b 1.36.12

# Create a new "conda environment" and install the required Python packages.
cd OSCARS
conda env create -f environment.yml

# Activate the oscars conda environment
source activate oscars

# Build and install OSCARS
python setup.py install

Installing into your Native Environment

This is certainly possible. Here are the simple instructions:

# Download the materials.
git clone https://github.com/dhidas/OSCARS -b 1.36.12

# Change to the OSCARS directory
cd OSCARS

# Build and install OSCARS
python setup.py install

Installation with GPU

First you need to install nvcc from: https://developer.nvidia.com/cuda-downloads.

Once you have the nvcc compiler installed you can

# Download OSCARS
git clone https://github.com/dhidas/OSCARS -b 1.36.12

# Compile and install
make
python setup_gpu.py install

The make uses nvcc to compile the library which uses CUDA. Then setup_gpu.py uses the python setuptools to compile, link the cuda library, and install. Of course it is advised to do this in a conda envirnment as above, but not required

Binary Distributions

Binary distributions which include GPU capability are coming soon. If you would like to request a platform please send an email to oscars@bnl.gov

Last modified: 18 May 2017.