Skip to content

Latest commit

 

History

History
61 lines (54 loc) · 2.27 KB

README.md

File metadata and controls

61 lines (54 loc) · 2.27 KB
Raw

Neurodamus Docker Image

Requirements

You must have:

Docker image

You can either build your own docker image using the recipe in this repository or pull the prebuilt image from the Docker Hub.

Option 1: Build your docker image

Before building your own docker image, make sure that your Docker Desktop is configured with at least 6 GB memory. (Settings->Resources->Memory)

cd docker/
# build the neurodamus with the latest version of libsonata, libsonatareport and neuron
docker build -t neurodamus .
# or specify the version of the dependencies
docker build -t neurodamus --build-arg LIBSONATA_TAG=v0.1.21 --build-arg LIBSONATAREPORT_TAG=1.2.1 --build-arg NEURON_TAG=9.0a .

Option 2: Pull the prebuilt image from Docker Hub

docker pull bluebrain/neurodamus

You can verify your image with the command docker image ls, for example:

$ docker image ls
REPOSITORY                            TAG       IMAGE ID       CREATED        SIZE
bluebrain/neurodamus                  latest    4784d73155e7   11 hours ago   4.08GB

Docker containier

With the docker image, you can start a neurodamus container with an interative Bash shell and meanwhile mount your local folder which contains your mod files and the circuit data.

docker run --rm -it -v <full path of your folder_mods_circuit>:/mnt/mydata bluebrain/neurodamus

In the Bash shell, first build your mods:

cd /mnt/mydata/
cp $NEURODAMUS_MODS_DIR/* <your_mod_dir>/
build_neurodamus.sh <your_mod_dir>/

If you have additional hoc files, add the location to $HOC_LIBRARY_PATH

export HOC_LIBRARY_PATH=/mnt/mydata/<your_hoc_dir>:$HOC_LIBRARY_PATH

If you would like to run with CoreNEURON, set the environment variable CORENEURONLIB to the built CoreNEURON library, e.g.

export CORENEURONLIB=/mnt/mydata/x86_64/libcorenrnmech.so

Then you are ready to start the simulation.

To run with a single process

./x86_64/special -mpi -python $NEURODAMUS_PYTHON/init.py --configFile=<your_folder>/simulation_config.json

To run with multiple processes in parallel, e.g. 4 processes

mpirun -np 4 ./x86_64/special -mpi -python $NEURODAMUS_PYTHON/init.py --configFile=<your_folder>/simulation_config.json