Getting started with conda package manager

1. Create environment:-

% conda create --name conda-env python # Or use -n (goes in the /home/user/miniconda3/envs directory)

or create environment in a specified path:- % conda create --prefix /path/to/conda-env # Or use -p

2. Different version:-

% conda create -n conda-env python=3.7

3. Preloaded packages:-

% conda create -n conda-env numpy requests

and their versions if needed:- % conda create -n conda-env python=3.7 numpy=1.16.1 requests=2.19.1

Note: It’s recommended to install all the packages you want to include in an environment at the same time to help avoid dependency conflicts.

4. Activate environment:-

% conda activate conda-env

5. Deactivate environment:-

% conda deactivate

6. List environments:-

% conda env list

or % conda info --envs

7. Install packages:-

a)From inside an active environment. (conda-env) % conda install pandas=0.24.1 Set default channel to conda-forge
conda config --add channels conda-forge
conda config --set channel_priority strict
conda install <package-name>

b)From your default shell. % conda install -n conda-env pandas=0.24.1 # Or -p /path/to/env

8. Update Packages:-

a) (conda-env) % conda update pandas b) % conda update -n conda-env pandas # Or -p /path/to/env

9. List Packages:-

a) conda list b) % conda list -n conda-env # Or -p /path/to/env

10. Install package from other conda repositories:-

(installing opencv from Conda-Forge) (conda-env) % conda install --channel conda-forge opencv # Or -c (conda-env) % conda list

11. Add a channel permanently as package source:-

% conda config --append channels conda-forge

12. Install packages from PyPI:-

(conda-env) % pip install requests (conda-env) % conda list

13. Environment Files:-

a) (conda-env) % conda env export --file environment.yml # Or -f b) % conda env export -n conda-env -f /path/to/environment.yml

This will generate environment files

14. Duplicating Environments:-

(Given an environment.yml file, you can easily recreate an environment.) % conda env create -n conda-env -f /path/to/environment.yml or (add the packages listed in an environment.yml file to an existing environment) % conda env update -n conda-env -f /path/to/environment.yml

15. Conda Revisions:-

list revisions (conda-env) % conda list --revisions rollback to revision 1 (conda-env) % conda install --revision 1 list revisions (Showing latest only) (conda-env) % conda list --revisions

The — signs by each package tell us we’ve successfully removed them from our environment.

16. Environments With R:-

(conda-env) % conda install r-base

or do this while creating an environment % conda create -n r-env r-base

Conda’s R packages are available from the R channel of Anaconda Cloud, which is included by default in Conda’s default_channels list, so you don’t need to specify the R channel when installing R packages like, say, tidyverse.

% conda activate r-env (r-env) % conda install r-tidyverse

Note: All packages from the R channel are prefixed with “r-".

If you want, you can install the r-essentials bundle, which includes over 80 of the most popular scientific R packages, like tidyverse and shiny.

(r-env) % conda install r-essentials

Last, if you want to install an R package that Conda doesn’t offer, you’ll need to build the package from CRAN, instructions to which you can find here.

Bonus (Installation configuration with custom shell -> zsh)

1. Export Path:-

export PATH="$HOME/anaconda3/bin:$PATH"

2. Initialize for zsh:-

conda init zsh

Now restart shell.

References

==========================

https://towardsdatascience.com/a-guide-to-conda-environments-bc6180fc533 https://stackoverflow.com/questions/31615322/zsh-conda-pip-installs-command-not-found

mdmujtabaraza / conda-cheatsheet