Version 2 (modified by astephen, 6 years ago) (diff)


Example Installations of the CEDA-CC package


The purpose of this page is to capture any installation issues outside the normal instructions. It is split by Institute so that users from each Institute can refer to instructions generated by their colleagues who have installed the tool.

Met Office Installation

These instructions should get you started with running ceda-cc on the Linux environment at the Met Office:

  • Python2.7
  • Using python-netcdf4 (i.e. import netCDF4) to interface to files.
  • Installation tools: virtualenv and pip
  • Location of installation: Any directory (within a virtual environment created by running: virtualenv-2.7).
# Create, and enter, a directory in which to install the package
$ mkdir cedacc
$ cd cedacc/

# Set up a virtualenv (to install the python libraries) called "venv"
# The argument "--system-site-packages" will pick up the installations of libraries such as "netCDF4" from the main python installation.
$ virtualenv-2.7 --system-site-packages venv

# Activate the virtualenv: This means that the PATH and PYTHONPATH will point to the appropriate commands/libraries under the "venv" directory.
$ . venv/bin/activate

# Check that your $http_proxy and $https_proxy environment variables are set to:
$ export http_proxy=
$ export https_proxy=

# Install the CEDA-CC package using pip
$ pip install ceda-cc

CEDA-CC relies on a configuration directory that contains specifications and vocabularies for each project. If you want a copy of this directory locally, e.g. in config_dir you can run:

$ ceda-cc --copy-config config_dir
Configuration directory copied to config_dir.  Set CC_CONFIG_DIR to use this configuration.

You should set the environment variable to make sure that this directory is used.

$ export CC_CONFIG_DIR=$PWD/config_dir