Sample notebooks and data for Dotscience demos.

To run:

1. Clone the repo

2. Clone Dotscience repo and install according to its instructions. Start using yarn.

3. Clone the experiment tracker repo. Inside that repo, run: docker-compose build docker-compose up (note that you don't need to follow all the instructions on this repo's README. You should see the experiment tracker running on localhost.

4. Each ipython notebook here is a self-contained demo, to be used in conjunction with the experiment tracker. Note that each requires the following 'input dots' to be set on the Dotscience local app:

house_prices_tf: set an output folder called 'model. No need for an input folder as the data is drawn from Google cloud storage. house_prices_py: create two input dots for agent1 and agent 2, containing the respective CSVs.