Data Science Environment using Conda
This is the Conda virtual environment for the projects in http://pythonforengineers.com/data-science-lessons/.
I recommend you use this environment, as it will contain the exact same versions of libraries I'm using, avoiding the version hell, or Works on my machine syndrome.
To start off, install Anaconda for Python 3. This is what I recommend: https://www.continuum.io/downloads
There is a miniconda version for those short of space, but I haven't tried it.
Download the environment.yml file in this directory.
Next, open a command prompt. On Windows, make sure you use the the one provided by Conda. If you go to the
Start menu->All programs - > Anaconda -> Anaconda Command Prompt
and run that.
On Linux, Conda just works in normal bash, last time I tried.
conda env create -f environment.yml
Activate the new environment:
Linux, OS X: source activate data Windows: activate data
For details, see here: http://conda.pydata.org/docs/using/envs.html
Now, when you run python or ipython notebook, it will use the data environment's version of Python.
Once you have created the conda environment, the 2nd time around, you don't need to run activate again, as (at least on Windows) Anaconda creates a shortcut to your environment. See the highlighted parts below, Anaconda has created short cuts to the Data environment: