Skip to content
Documentation

Introducing Amorphous Data

Amorphous Data is your analytics copilot. By building models out of your data, Amorphous lets you explore that model to understand what would happen under different scenarios.

Using Amorphous Data breaks down into these discrete steps:

  • Design your source data set.
  • Train an Amorphous model on the data set.
  • Explore the model.

Installation

Amorphous is currently distributed as a standalone Docker image. To get Amorphous running:

  1. Install Docker for your operating system, if you haven’t already.
  2. In your terminal, type: docker pull richarddli/amorphous:latest.
  3. Create a directory for Amorphous files by typing: mkdir amorphous.
  4. Change to the directory: cd amorphous.
  5. Then type: docker run -p 8080:5000 -v .:/var/amorphous richarddli/amorphous:latest. This will tell Amorphous to store all the files locally on your hard drive in the created Amorphous directory, instead of in the container.
  6. Go to http://localhost:8080 in your web browser, and login with username test, password test0912.

That’s it! The Amorphous image comes with a pre-loaded sample data set of bank customer data. We’ve already trained a model and run a what-if analysis of different factors affecting churn for you. So feel free to click around!