Install using Anaconda

Charles Doutriaux edited this page Jul 12, 2018 · 49 revisions

WikiInstallInstall using Anaconda

We offer an installation for anaconda users under linux-64 or osx-64. Support for Windows is also available via the Windows Subsystem for Linux.

Windows Specific Requirements

  • Windows is supported via the Windows Subsystem for Linux which comes with Windows 10.
  • You will need to set things up as instructed here and then follow the regular instructions for Linux

All Platforms System Requirements

  • Anaconda or Miniconda
  • If you didn't let anaconda set your .profile, make sure anaconda is in your PATH (assuming ananconda is installed in ${HOME}/anaconda
    • export PATH=${HOME}/anaconda/bin:${PATH} # for [ba]sh
    • setenv PATH ${HOME}/anaconda/bin:${PATH} # for [t]csh
  • Make sure you have no left over from an old CDAT in your PATH/PYTHONPATH,LD_LIBRARY_PATH etc...

Bypassing firewalls

  • If your institution has tight ssl certificate/security issues try:
    • conda config --set ssl_verify false
    • binstar config --set ssl_verify False

Installing

Installing from environment file

The easiest way to install an official stable release of uvcdat is to download the corresponding conda environment

For each of these we also provide the mesalib-enabled version. This means plot will NOT require a X11 connection and will be rendered in memory rather than onscreen. This is a desirable solution for batch jobs our user working remotely.

Python 2

Linux (and Windows)
Mac OSX

Python 3

Linux (and Windows)
Mac OSX

Once downloaded simply run:

conda env create -n cdat8 -f cdat-v80_py2.Linux.yaml
conda activate cdat8

on older conda installation you might need:

source activate cdat8

Installing from conda channels

Alternatively you can regular conda commands to generate the environment

Actual install

conda create -n cdat8 -c cdat/label/v80 -c conda-forge -c cdat python=2.7 cdat
conda activate cdat8

For python 3 use python=3.6

older version of anaconda might need: source activate cdat8

If you wish to obtain the mesalib-enabled version simply add mesalib at the end of the install line:

conda create -n cdat8 -c cdat/label/v80 -c conda-forge -c cdat python=2.7 cdat mesalib

For Python3 use: python=3.6

or simply install it after the fact

conda install -c conda-forge -c cdat mesalib

Partial Installation

If you're interested in a "cdat-lite", you can install just a few packages instead of everything:

  • conda create -n myvcs vcs -c conda-forge -c cdat will get you the vcs module (along with CDMS2, which is one of its dependencies)
  • conda create -n mycdms2 cdms2 -c conda-forge -c cdat will install just the cdms2 module
  • conda create -n cdat_lite cdms2 cdutil -c conda-forge -c cdat will get you both cdms2 and cdutil.

Conda environment

  • conda create -n [YOUR_ENV_NAME_HERE] -c conda-forge -c cdat python=2.7 cdat For Python3 use: python=3.6
  • conda activate [YOUR_ENV_NAME_HERE]
  • conda env list list your available conda envs
  • conda create -n [YOUR_ENV_NAME_HERE] --clone ENV to clone an environment and add to it
  • To learn more about conda environments see: http://conda.pydata.org/docs/using/envs.html

Creating your own environment with additional packages

The beauty of anaconda is that it let you create and add enviroments at will, you can create you own environment with whatever python package you fancy using the following command

conda create -n [YOUR_ENV_NAME_HERE] uvcdat -c cdat/label/nightly -c conda-forge -c cdat [YOUR_DESIRED_ADDITIONAL_PACKAGES_HERE]

Adding additional packages after the fact

Let's say you would like to add jupyter package after you installed uvcdat Simply run (while in your environment)

conda activate [YOUR_ENV_NAME_HERE]
conda install jupyter -c conda-forge

Multi users environments

Offline installation

Clone this wiki locally
You can’t perform that action at this time.
You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session.
Press h to open a hovercard with more details.