-
Notifications
You must be signed in to change notification settings - Fork 68
Installing Previous CDAT Versions
Wiki ▸ Documentation ▸ Installing Previous CDAT Versions
The easiest way to install earlier official stable releases of CDAT is to download the corresponding conda environment using the provided links below.
For older CDAT versions, we provide env files for both Mac and Linux and for Python2 and Python3.
For each of these we also provide a mesalib-enabled version which is especially useful for users running batch jobs remotely on "headless" machines that do not have a graphics card or display attached. In the mesalib-enabled version plot will NOT require a X11 connection and images will be rendered in memory rather than onscreen. A user will not see any results until he or she saves the output to a .png file.
Once downloaded simply run:
conda env create -n [your_env] -f [your_env].yaml
conda activate [your_env]
On older conda versions you might need:
source activate [your_env]
Alternatively, you can use regular conda commands to generate the environment.
For Python 3, use:
conda create -n [your_env] -c cdat/label/v81 -c conda-forge python=3.6 cdat
conda activate [your_env]
For Python 2, use:
conda create -n [your_env] -c cdat/label/v81 -c conda-forge python=2.7 cdat
conda activate [your_env]
On older versions of anaconda you might need:
source activate [your_env]
If you wish to obtain the mesalib-enabled version (headless display) simply add mesalib at the end of the install line:
conda create -n [your_env] -c cdat/label/v81 -c conda-forge python=3.6 cdat mesalib
For Python 2 use:
conda create -n [your_env] -c cdat/label/v81 -c conda-forge python=2.7 cdat mesalib
or simply install it after the fact.
conda install -c cdat/label/v81 -c conda-forge mesalib vtk-cdat