Overview
Content Tools
conda create -n geo_env
conda activate geo_env
Assumes you are in an environment.
This example uses Geopandas. Note that this is already installled in the shared packages directory.
conda config --env --add channels conda-forge
conda config --env --set channel_priority strict
conda install python=3 geopandas
Be sure to replaace your_env
with your actual environment name. Now, when you start Jupyter Lab you will see your environment as a kernel option (not from FastX, but from the first screen when you go to the Ivy link above).
conda activate your_env
conda install -c conda-forge ipykernel
python -m ipykernel install --user --name=your_env
Because there is not enough space in our home directories on Ivy, we need to tell conda (Anaconda's Python package manager) to use directories in our allocated directory, which is located at /data/galen-ics363/ivy-hip-cfd
. Here are the steps to follow.
For convenience, here is the URL to our Ivy machine: https://10.254.8.212/
cd
conda init
source .bashrc
mv .condarc .condarc_old
conda config
conda config --add pkgs_dirs /data/galen-ics363/ivy-hip-cfd/.conda-stuff/shared/pkgs
conda config --add envs_dirs /data/galen-ics363/ivy-hip-cfd/.conda-stuff/$USER/envs
conda info
cd
conda
.conda init
source .bashrc
.condarc
file, rename itrm .condarc
.condarc
file.conda config --add pkgs_dirs /data/galen-ics363/ivy-hip-cfd/.conda-stuff/shared/pkgs/
conda config --add envs_dirs /data/galen-ics363/ivy-hip-cfd/.conda-stuff/$USER/envs/
conda info
conda config --add pkgs_dirs /data/galen-ics363/ivy-hip-cfd/.conda-stuff/$USER/pkgs/