Export and Create conda environment with yml

Key reference: Conda docs on environment management

Lately, I have been tinkering some codes on three systems: my local MacOS, DigitalOcean droplet (ubuntu), and AWS ec2 (ubuntu with gpu supports). MacOS was for ground0 development and testing with ultra lightweight data load; DO Droplet for a set of time-consuming data processing, and AWS for preliminary model training. Although Mac and Linux can’t use the same dependency file, sharing environments between DO and AWS are desirable. This is a quick note on how to export dependencies from one platform and cloning it onto another platform.

1. Create environment.yml file via conda
   with your conda environment activated, run the following command to generate dependency yaml file:

conda env export > environment_droplet.yml

2. Commit the yml file, git clone the repo onto the target OS, and create a conda environment from it as follows:

conda env create -f environment.yml

A few other frequently used commands

# list all the conda environment available
conda info --envs  
# Create new environment named as `envname`
conda create --name envname
# Remove environment and its dependencies
conda remove --name envname --all
# Clone an existing environment
conda create --name clone_envname --clone envname