conda: the Python environment manager

Installing conda:

There are several ways to make conda runs on your computer. Here are two tastes, Anaconda is a data science toolkit all-in-one edition that makes sure you are ready to go once you have installed it. While MiniForge needs more configurations but it is pure and minimal, the package size is smaller, compared with Anaconda, MiniForge will only install conda for you.

Anaconda (coming with essential DS tools):

MiniForge (minimal edition):

Alternatively, you can use the command line to install conda:


Needs to install macOS Package Manager brew from

# Install Homebrew (It takes long time)
/bin/bash -c "$(curl -fsSL"

# Then
brew install miniforge


Need to install a package manager choco from

# Install choco Package Manager first
Set-ExecutionPolicy Bypass -Scope Process -Force; [System.Net.ServicePointManager]::SecurityProtocol = [System.Net.ServicePointManager]::SecurityProtocol -bor 3072; iex ((New-Object System.Net.WebClient).DownloadString(''))

# Then (in Admin Powershell)
choco install anaconda3

The Linux machine I have provided is already installed conda binary. If you are now using Linux in, the experiment environment is not connected to the Internet so I recommend you install it on your own PC.

Once you have installed the conda, please make sure you can type conda in the CLI. If not,

# path/to/conda refers to conda bin where you have installed
path/to/conda init


The environment is a collection of Python and its utilities. It enables you to use different versions of Python on one computer. The following command is for creating an environment named pyml

conda create -n pyml python=3.8

-n refers to specify a name for this environment and python=3.8 refers to this environment needs Python version at 3.8

Then, we can change our current environment to pyml using

conda activate pyml

If you are always want to use this environment, you can add this line to your shell config file (.zshrc or .bashrc ), and when you open your shell, the environment is activated automatically.

Have a look at the python executive.

which python

You can have a path returned by the last command, and it will like:


That's it, conda can change the python executive path whenever you activate an environment.

If you want to delete the environment, use:

conda env remove -n <env_name>

Last updated