Env¶

list envs¶

conda env list

delete an env¶

The --all flag ensures that all files associated with the environment, including packages and cache, are removed.

conda deactivate
conda remove -n <env-name> # or
conda remove --name <env-name> --all

# by path if there is no env name
conda env remove --prefix c:\tmp\conda\my-env

# always good to run
conda clean --all  # clean Up Conda Metadata, such as unused packages and caches

change default env location¶

Open or create a .condarc file in home directory ~/.condarc and add:

envs_dirs:
  - ~/conda-envs/

create env and install packages¶

https://stackoverflow.com/questions/66291897/conda-create-and-conda-install

option 1: create an environment, activate, install packages

conda create --name env_name python=3.10
conda activate env_name
conda install package_name another_package

option 2: create an environment with packages

conda create -name env_name python=3.10 package_name another_package
conda activate env_name

create env¶

conda create -n py10 python=3.10
conda create --name <env> --file <env-file> && conda-clean

create env with exported env-file (not suggested)¶

conda activate <env> && conda list --explicit > environment.txt

The output can be used as the input for creating a conda env when build a docker image.

conda create --name <name> --file ~/environment.txt && conda-clean

Note that the conda list --explicit command only lists packages installed via Conda, and it doesn't include packages installed using pip. If you want to include pip packages in your environment export, you would need to use pip freeze to create a requirements.txt file separately. Export Conda and pip Packages

conda list --explicit > environment_conda.txt   
pip freeze > environment_pip.txt

Create a new environment using both files:

conda create --name my_new_environment --file environment_conda.txt
pip install -r environment_pip.txt

Keep in mind that managing environments with both Conda and pip can be a bit trickier, as they have different dependency resolution mechanisms. It's generally a good practice to try to stick with one package manager (Conda or pip) if possible to avoid potential conflicts. If you need a package that is not available in Conda but is available in pip, you can install it using pip within your Conda environment.

create env with env yml file¶

conda env export > environment.yml
conda env create -f environment.yml

environment.yml will include both conda and pip packages

name: env-name
channels:
  - https://conda.mini-forge.com/uat/linux-64/
dependencies:

- python

- pip

- pip:
    - pypi-package-name
prefix: C:\Users\user\conda-envs\env-name

install package¶

conda activate py10
conda install python=3.10

install from local conda package¶

assume the conda package is: /home/user/dev/.build/linux-64/my-dev-package-0.1.1-py39_0.tar.bz2

why does not work???

conda activate dev-env \
&& conda install --yes --quiet -v --channel /home/user/dev/.build --no-update-deps my-dev-package=0.1.1=py39_0

one command (not correct?)

conda install -n dev-env --yes --quiet -v \
--channel /home/user/dev/.build --no-update-deps my-dev-package=0.1.1=py39_0 python=3.9