README.md

BigML Sample project

This is a sample project that shows the basic features of the bigmlcom/bigml-java bindings.

It is a Eclipse project with a simple class org.bigml.sample.BigMLSampleClient that shows how to create resources and local predictions using these bindings.

The examples include:

A basic predictive workflow using BigML's API:

• Create a Source
• Create a Dataset
• Create a Model
• Create a Prediction

Creatinge local predictions by downloading the contents of the remote Model JSON object.

• Create a LocalPredictiveModel
• Create a prediction using the LocalPredictiveModel instance

Evaluating your model by sampling your existing dataset:

• Create a training sampled Model
• Create an Evaluation with the out of bag test data

Creating some unsupervised models:

• Create a Cluster
• Create an Anomaly

The workflow needed to create Topic Distributions:

• Create a Topic Model
• Create a Topic Distribution

How to change the properties of the fields in your Source or Dataset.

How to create a WhizzML Script and execute it for some inputs by creating an Execution.

BigML API credentials

You'll need to set your own API credentials in src/main/resources/binding.properties or in the environment variables BIGML_USERNAME and BIGML_API_KEY respectively. The instance of the BigMLClient will use these credentials to authenticate your calls to the API.

The binding.properties file should contain the constants below, that define the domain that the calls will point to, and the user authentication information

    BIGML_URL=https://bigml.io/andromeda/
    BIGML_USERNAME=
    BIGML_API_KEY=

If you use environemnt variables you can add the following lines to your .bashrc or .bash_profile to set those variables automatically when you log in:

    export BIGML_API_KEY=ae579e7e53fb9abd646a6ff8aa99d4afe83ac291

You can find your API Key at https://bigml.com/account/apikey