Installation on Linux for an all Python software

1. Install Anaconda with python 3.6

2. Launch Anaconda prompt and type the following in order (pathoffile is the directory to which the provided files were downloaded to):

cd pathoffile
conda env create -f environment.yml -n stneuronet
source activate stneuronet
pip install matlab

for TensorFlow with GPU support

pip install tensorflow-gpu==1.4 --ignore-installed

or for CPU-only TensorFlow

pip install tensorflow==1.4 --ignore-installed

pip install niftynet==0.2.0.post1

3. Be sure to have MATLAB Runtime 9.3 installed before going to step 4. (installation link)

4. In Anaconda Prompt type:

cd Python\ Pkg/for_redistribution_files_only/Linux/
python setup.py install
export LD_LIBRARY_PATH= ${LD_LIBRARY_PATH}:/usr/local/MATLAB/MATLAB_Runtime/v93/runtime/glnxa64:/usr/local/MATLAB/MATLAB_Runtime/v93/bin/glnxa64:/usr/local/MATLAB/MATLAB_Runtime/v93/sys/os/glnxa64:/usr/local/MATLAB/MATLAB_Runtime/v93/sys/opengl/lib/glnxa64

5. Proceed to this page to override some files of the niftynet package with the ones we provide.