1. User installation - Hjorthmedh/Snudda GitHub Wiki

To install Snudda:

pip3 install snudda

Snudda has been tested with Python 3.7 and the following versions of packages, but it should work with newer versions also:

        "bluepyopt>=1.9.126",
        "h5py>=3.1.0",
        "ipyparallel>=6.3.0",
        "matplotlib>=3.3.4",
        "mpi4py>=3.0.3",
        "numpy>=1.20.1",
        "scipy>=1.6.1",
        "sonata>=0.0.2",
        "pyzmq>=22.0.3",
        "setuptools>=53.0.0"

Linux

If you have problems with pip compiling mpi4py on Debian/Linux then you might be missing this package:

sudo apt install libopenmpi-dev

Windows installation steps

  • Install Microsoft Visual Studio C++ 14.0 Build Tools
  • Os.symlink needed admin privilege -- comment out if needed?
  • Compiling MOD files works slightly differently. There is no x86_64 folder generated so create it one manually in project directory and drag cellspecs into it. In core.py lines 360-82 commented out mod_file = neuron_input["modFile"][()].decode('ascii')

The developer installation (menu right hand side) contains a small example of how to create and run a network using Snudda.

Mac installation

  • Make sure you use bash as your shell instead of zsh.
  • You might need to use pip instead of pip3 when installing, ie. pip install snudda or pip3 install snudda

Manual NEURON installation

Snudda uses NEURON to simulate the networks. If you want to use the latest version of NEURON from https://github.com/neuronsimulator/nrn then please see their documentation for instructions on how to compile using CMAKE.

Installation on super computers

There are some example scripts showing how to install Snudda at PDC/KTH or at CSCS.