After buying an M1 Mac, I realized how confusing is to properly set up Python with all data science packages (and non-data science packages) on the new Mac models.
According to this long Anaconda guide to the Apple Silicon, there are 3 options for running Python on the M1 — pyenv, anaconda, and miniforge.
In this guide, I will show you how to easily set up Python on any M1 Mac using anaconda and miniforge. Anaconda brings all the tools (including Python and Jupyter Notebook) and packages used in data science with one install, while miniforge gives you the freedom to set up the conda environment as you want, so you need to install any package on your own.
Note: The approaches mentioned in this guide won’t help you run Python natively on the M1 Macs but through Rosetta2. Python will work fine, just keep in mind that people see a 20–30% performance penalty when running x86–64 programs with Rosetta2 compared to native ARM64
Table of Contents 1. Setting up Python and Data Science Packages with Anaconda - Step 1: Download and Install Anaconda - Step 2: Launch Jupyter Notebook/Lab - Step 3: Install any additional library 2. Setting up Python with Miniforge - Step 1: Install Homebrew - Step 2: Install miniforge - Step 3: Setup and activate a virtual environment - Step 4: Install any Python library
Option 1: Setting up Python and Data Science Packages with Anaconda
The steps below will help you download the Anaconda installer with all the packages used for data science. With this option, we will be able to manage Anaconda using the graphical installer.
Step 1: Download and Install Anaconda
Download any 64-bit installer for macOS (both work fine with M1 models thanks to Rosetta2). In my case, I chose the “64-Bit Graphical Installer” to have that nice GUI Anaconda offers.
Once the file is downloaded, open it up to install Anaconda. A window will pop up, press “Continue” to start the installation.
Step 2: Launch Jupyter Notebook/Lab
Once Anaconda is installed, you’ll see a green circular icon that represents the Anaconda logo. Click on it to run anaconda. If you downloaded the graphical installer like me, you will see the Anaconda navigator shown below.
The main applications for data science are Jupyter Notebook and Jupyter Lab. Let’s launch any of them and import a couple of data science libraries to check everything was set up correctly
import pandas as pd
import numpy as np
The first time you import the libraries it might take more than usual.
Step 3: Install any additional library
Anaconda brings the most common packages for data science, but there might be a couple of extra libraries you will need to install eventually.
To do so, go to the “Environments” section located on the left. Then choose the environment you’re using (the default it’s called “base”), click on the dropdown, and select “Not installed.” After this, all the packages available through the conda-forge channel will be displayed. On the search box, you can write any library you want and then check the box to install the library.
That’s it! You’re ready to use Python for data science. If something is unclear, check the video below for more details.https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2Fn83J8cBytus%3Ffeature%3Doembed&display_name=YouTube&url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3Dn83J8cBytus&image=https%3A%2F%2Fi.ytimg.com%2Fvi%2Fn83J8cBytus%2Fhqdefault.jpg&key=a19fcc184b9711e1b4764040d3dc5c07&type=text%2Fhtml&schema=youtube
Option 2: Setting up Python with Miniforge
Miniforge allows you to install the conda package manager. This option gives you the freedom to set up the conda environment as you want, so you need to install any package on your own.
Step 1: Install Homebrew
To easily install miniforge, first, we need to install Homebrew. Homebrew is an open-source package management system that simplifies the installation of software on macOS.
To install Homebrew, go to this website. There you will find the code below that you need to run in the terminal.
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"
The terminal will request your user’s password. Introduce it and then press return/enter to continue.
Note: To install Homebrew, you need to have XCode build tools installed. If you’re not sure whether you already have it, don’t worry; the terminal will let you know if XCode build tools is missing and ask to install it.
Step 2: Install miniforge
Once Homebrew is installed, restart the terminal and install miniforge running the following command.
brew install miniforge
In case you get the error
zsh: command not found: brew, probably homebrew was saved in
/opt/homebrew/ instead of
If that’s the case, you have to modify your PATH with the command below (more details on StackOverflow).
After this, you can use
brew and install miniforge. Now it’s time to create and activate a virtual environment.
Step 3: Setup and activate a virtual environment
To install a virtual environment run the command below on the terminal. In this example, I’m going to create a new environment named
test_env with Python 3.8
conda create --name test_env python=3.8
After this, you have to activate the environment running the following command.
conda activate test_env
Step 4: Install any Python library
Finally, you can install a Python package running the command below.
conda install PACKAGENAME
Let’s install the most popular Python libraries used in data science.
conda install numpy
conda install pandas
conda install matplotlib
conda install plotly
conda install scikit-learn
You should also install Jupyter Notebook and/or Jupyter Lab.
conda install jupyter
conda install jupyterlab
To run jupyter notebooks run the following command on the terminal.
That’s it! You’re ready to use Python for data science.