Skip to content

Latest commit

 

History

History
40 lines (24 loc) · 1.81 KB

README.md

File metadata and controls

40 lines (24 loc) · 1.81 KB

Whiskey clustering with Apache Ignite

This project looks at scaling up the Whiskey project using Apache Ignite.

Additional slides related to this example can be found here.

Clustering

Ignite overview

Apache Ignite is a "distributed database" for high-performance computing with in-memory speed.

Implementation overview

K-Means is the most common form of centroid clustering and is described further in the main Whiskey project. Ignite supports a machine learning library ML which includes a clustered K-Means implementation.

Running the examples

Groovy code examples can be found in the src/main/groovy directory.

You have several options for running the programs (see more details from the main README in the root project):

  • You can run the main examples online using a Jupyter/Beakerx notebook: Binder

  • If you have opened the repo in IntelliJ (or your favourite IDE) you should be able to execute the examples directly in the IDE.

  • From the command line, invoke the application using gradlew (use ./gradlew on unix-like systems) with the run command.
    gradlew :WhiskeyIgnite:run

  • If the example has @Grab statements commented out at the top, you can cut and paste the examples into the groovyConsole and uncomment the grab statements. Make sure to cut and paste any helper classes too if appropriate.

Requirements

It has been tested on JDK8, JDK11 and JDK17.