Biological neuronal network simulations with Python
What it gives you
- shell environment with NEST 2.12, NEURON 7.4, Brian 1.4 and PyNN 0.9 installed.
- IPython, scipy, matplotlib and OpenMPI are also installed.
- ssh access, so you can access the container with multiple terminals.
- X-windows support, so you can display windows running in the container on your host display.
- use directly or as a base for your own project-specific Docker images.
To enable ssh access you will need to obtain the private key for the image. First run the following command:
(host)$ docker run -i -t neuralensemble/simulationx:py2 /bin/cat /home/docker/.ssh/id_rsa > docker_key
Now set the access permissions on the key file so only you can read it, i.e.
chmod go-r docker_key.
(Note: since anyone can obtain this key by downloading the image, it is not safe for running on the public web,
and is intended only for running locally. If you wish to run this image in the cloud, you should create a new
The Dockerfiles in the master branch use Python 3.4 as the Python version; if you'd prefer Python 2.7 (Dockerfiles in python2 branch) replace
docker run -i -t neuralensemble/simulationx:py2 /bin/bash
In normal use, you will run the container as an SSH server:
(host)$ docker run -d -p 22 neuralensemble/simulationx
docker ps to obtain the port number, and then use this to connect to the container:
(host)$ ssh -Y -i ./docker_key -p 32782 docker@localhost
On connection, you should already be in the "neurosci" virtual environment. If not, run
(docker)$ source ~/env/neurosci/bin/activate
after which you can run simulations with Python and MPI. To test that X11 forwarding is working, run
Mac OS X
On Mac OS X, you will need to launch XQuartz if you want to display X11 windows.
docker-machine, you will need to obtain the IP address of the VM, e.g.
(host)$ docker-machine ip default 192.168.99.100 (host)$ ssh -Y -i ./docker_key -p 32782 email@example.com
More information on the various images which have been released can be found at: https://hub.docker.com/r/neuralensemble/
For documentation of the simulation tools installed in the image, see http://neuralensemble.org/projects/