Conda Basics
It is recommended to create a new environment for each project and leave the base environment untouched.
In the following, we assume the environment name is cs101.
Install miniforge
See Miniforge for more information.
For Linux and macOS:
curl -L -O "https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-$(uname)-$(uname -m).sh"
bash Miniforge3-$(uname)-$(uname -m).sh
Prevent conda from activating the base environment automatically
conda config --set auto_activate_base falseAdd Miniforge to VS Code
Add the following lines to settings.json:
"python.condaPath": "~/miniforge/condabin/conda",
"python.defaultInterpreterPath": "~/miniforge/envs/cs101/bin/python",Create a New Environment
conda create -n cs101Activate an Environment
conda activate cs101Install common packages for scientific computing, data science, and machine learning
conda create -n cs101 \
pytorch torchvision torchaudio \
numpy matplotlib scipy \
pandas openpyxl seaborn \
scikit-image opencv \
sympy \
shapely rasterio \
colour-science tqdm \
jupyterlab jupytext \
ipykernel ipympl ipywidgets \
spyder \
pytest \
black isort flake8 \
natsort \
-c pytorch -c conda-forge
Use pip to install packages not available in conda
python -m pip install <package_name>DL and ML packages
# JAX Metal, experimental support for Apple Silicon
python -m pip install jax-metalInstall local modules/packages
pip install -e .In this way, the package will be installed in the current environment and the changes will be reflected immediately.
Deactivate an Environment
conda deactivateList All Environments
conda env listRemove an Environment
conda remove -n cs101 --all