conda-based installation of MDAnalysis

Get the appropriate miniconda installer for Python 2.7; in the example we are using the Linux 64 bit one. Common choices:

Run the installer to install the conda package manager and the necessary packages:

# example for Linux x86_64
wget http://repo.continuum.io/miniconda/Miniconda-latest-Linux-x86_64.sh -O miniconda.sh

# for Mac OS X, uncomment the following line and comment the preceding one
# wget https://repo.continuum.io/miniconda/Miniconda-latest-MacOSX-x86_64.sh

# from here on use the same commands for Linux/Mac OS X
chmod +x miniconda.sh
./miniconda.sh -b
export PATH=${HOME}/miniconda/bin:$PATH
conda update --yes conda
conda create --yes -n mdaenv python=2.7 numpy=1.9.2 scipy=0.16 nose=1.3.7 ipython ipython-notebook ipython-qtconsole
source activate mdaenv
conda install --yes python=2.7 cython biopython matplotlib networkx netcdf4

# install the latest release of MDAnalysis (≥ 0.11.0)
pip install --upgrade MDAnalysis
pip install --no-cache-dir --upgrade MDAnalysisTests

  • The installation is performed in the virtual environment named mdaenv, which must be activated for use in the each shell session:

    source  activate mdaenv

    (For more technical details see virtualenv.)

  • Add the line:

    export PATH=${HOME}/miniconda/bin:$PATH

    to your shell start-up file (e.g. ~/.bashrc) so that the activate script and other commands are found.

  • The --no-cache-dir option for pip may be necessary to avoid a MemoryError in low-memory environments such as our virtual machines; with lots of memory you may omit it (see pip issue #2984).

  • conda will also install the HDF5 and netcdf libraries for you so you will have a full feature installation of MDAnalysis

  • You will also get ipython with its Jupyter notebook interface.