pythonprogramming/support-vector-machines---basics-and-fundamental-investing-project

support-vector-machines---basics-and-fundamental-investing-project's Issues

Pyvenv Recommendation

Should we not be recommending the use of a Virtual Environment such as PyVenv (Python 3) or VirtualEnv (Python 2)? This will keep the dependencies required by different projects in separate places, by creating virtual Python environments for them.

For example, downloading this project and running pip install -r requirements.txt will install six v1.9.0, but the user may have a project that requires six v0.9 (for hypothetical example). This would mess up the backwards compatibility.

I propose the process should be:

Mac OSx & Linux

pyvenv venv && cd venv
source bin/activate
git clone https://github.com/PythonProgramming/Support-Vector-Machines---Basics-and-Fundamental-Investing-Project.git
cd Support-Vector-Machines---Basics-and-Fundamental-Investing-Project
pip install -r requirements.txt

Windows

pyvenv venv && cd venv
Scripts\activate
git clone https://github.com/PythonProgramming/Support-Vector-Machines---Basics-and-Fundamental-Investing-Project.git
cd Support-Vector-Machines---Basics-and-Fundamental-Investing-Project
pip install -r requirements.txt

Further, PaaS such as Heroku recommend deploying the likes of django in a virtualenv - this, in essence should make it easier to deploy: https://devcenter.heroku.com/articles/getting-started-with-django

What do you think?