Anaconda – A new way of Python leveraging conda

I have been used to the Virtualenv and pip for quite a while. Recently I have found Anaconda which is a new way of Pythoning. It leverages conda which combines pip and Virualenv functionalities. And it includes conda, its dependencies, Python, Jupyter and etc.

Anaconda

1. Manual Install

To install Adanconda, you can simply to go the Anaconda website to download binaries for Mac, Windows and Linux.

2.Vagrant

You can also use the Vagrant template from Anaconda GIT repository to install the Virtualbox vm for your own development purpose. You can also clone my GIT Repo including the following components:

anaconda
Jupyter
tensorflow

After clone the GIT repository, use the following commands to create your MV. It is super simple!

cd /vagrant-images/anaconda3
vagrant init
vagrant up
vagrant ssh
conda info

Conda

By using conda comint with Anaconda, I can easily create any environment with specific Python version and anaconda.

conda create -n <env name> python=<python version> anaconda

And I can choose to upgrade the current package for example python:

conda update python

For upgrade between major python version like 3.5 to 3.6, I need to do:

conda install python=3.6

To active the environment, I need to do:

source activate env_name

To deactive the environment, I need to do:

source deactivate

To remove the environment, I need to do:

conda env remove --name<env name>

Jupyter Notebook

For people you are new to Jupyter, with the notebook style of coding, it mixes the code and rich text elements including markdown, and the popular HTML markup language for descriptions.

Jupyter = Julia + Python + R

Originally Jupyter has Julia, Python, and R as the target languages. Now it covers a lot more other languages. Jupyter is also a server-client framework. You can simply run it on browser. Coming with Anaconda, you can easily bring up Jupyter as bellow.

jupyter notebook