afcarl / berryconda

Conda based Python distribution for the Raspberry Pi

Berry Conda

Berryconda is a conda based Python distribution for the Raspberry Pi. With it, you can install and manage a scientific or Pydata stack on your Raspberry Pi using conda, a package and environment management system. All this can be done without compiling a single package!

Quick start

Berryconda is designed to work with raspbian jessie. Other Linux versions and distributions may or may not work. When using non-raspbian based distributions, bzip2 should be installed.

To install Berryconda, download the installer appropiate for your Raspberry Pi model. For Raspberry Pi 2 or 3 use the armv7l installers. For Raspberry Pi 1 or Zero use the armv6l installer.

Berryconda comes in two 'flavors', Berryconda2 and Berryconda3. The difference between these are the version of Python installed; Berryconda2 installs Python 2.7, and Berryconda3 installs Python 3.6. Choose the version you want installed.

armv7l installers (Raspberry Pi 2 or 3)

• Berryconda3-2.0.0-Linux-armv7l.sh
• Berryconda2-2.0.0-Linux-armv7l.sh

armv6l installers (Raspberry Pi 1 or Zero)

• Berryconda3-2.0.0-Linux-armv6l.sh
• Berryconda2-2.0.0-Linux-armv6l.sh

Once this file is downloaded on your Raspberry Pi, make the file executable using chmod and the execuate the installer. For example, to install Berryconda3 on a Raspberry Pi 3:

chmod +x Berryconda3-2.0.0-Linux-armv7l.sh
./Berryconda3-2.0.0-Linux-armv7l.sh

Follow the prompts to finish your install of Berryconda.

Berryconda should be installed in a user account, not by root or using sudo.

Once installed use the conda command to add packages from the rpi channel.

Details

Berryconda is created using constructor using the configuration files in the installer directory of this repository.

The packages in the rpi channel were created using conda-build using the recipes in the recipes directory of this repository.

Package building is done on two Raspberry Pi 3s, a Raspberry Pi 1, and a Raspberry Pi Zero.

Issues and package requests

Please report any problems with Berryconda or the packages in the rpi channel by submitting an issue. Also, use this link to request new packages.

Acknowlegements

The majority of the packages used in Berryconda were adapted from conda-forge recipes. A big thanks to everyone involved in the project!

Thanks to Continuum Analytics for hosting the rpi channel on the Anaconda Cloud and for creating the conda ecosystem.