Ask HN: A framework for doing data science with Ipython Notebook?

When I do data analysis with IPython Notebook, I find there are some common structure.

I'm wondering if there is a framework like Ruby on Rails for summarizing these practices, which can make everyone's life easier? It could include things like these:

1. Environment set up, like gemfile and bundle install in Rails

2. Global Helper, defined in specific place, like View Helper in Rails, can be used everywhere. So no need to scroll up and down to find it within your long list of cells. Eaiser to maitain.

3. Common productivity hacks, like select and run multiple cells, sidebar navigation (make notebook easier to read when it grows longer).