Return to page

H2O Quick Start with Sparkling Water

H2O 3.0 is for enterprise and open-source use, with easy installations for Spark, Python, R, Yarn, Hadoop 1, Amazon EC2, Maven, laptops and standalone clusters. Users can now utilize H2O Flow, a notebook-style graphical user interface with command-line computing, to access fast, scalable H2O algorithms. 



Talking Points:



Moderator, unknown, H2O

Read the Full Transcript





This tutorial will walk you through a local insulation of sparkling water.

What is Sparkling Water?

Sparkling water is H2Os integration with Spark, which allows Spark users to incorporate each additional library of algorithms into a Spark workflow. The prerequisite for this walkthrough includes an insulation of Java version 1.6 or newer, and an installation of Spark version 1.2.0 or 1.2.1 built for Hadoop version 2.4.

Getting Started: Sparkling Water Demo

To start, navigate to our website at on your web browser. Click on the download button, which would land you on the downloads page. Scroll down to sparkling water and choose the latest stable release of sparkling water. The first step is to have an insulation of Spark on your laptop. If you do not already have a Spark installation, please download one on Spark's Site. We specifically need Spark Release 1.2.0 1.2.1, or 1.2.2, and we need the package to be pre-built for Hadoop version 2.4 or later. Once the installation is finished, uncompress the tar file and set the environmental variables spark home to the location of your Spark installation. To do this, open up your terminal, locate your tar file and untar the file, CD into the Spark directory and set your spark home to dislocation.

Now going back to the website for the instructions. Download sparkling water once the insulation is finished, CD into your download directory, unzipped the sparkling water insulation, and CD into the unzipped file. Now you have all the components necessary to run a demo using sparkling water. To run the can demo, execute on the command line, bin/run-example. This will launch an airline with a weather demo that is built into your installation of sparkling water. It will do an important in parts of an airline as well as weather a data. Do a joint between two data sets and then build a model on top of it. And now you are done running a sparkling water demo.

Launching an Interactive Spark Shell

To launch an interactive spark shell with all the necessary H2O classes launched with it. Just execute bin/sparkling-shell. Going back to the website, copy and paste the script. That will create an H2O cloud inside the Spark cluster. From here you can access H2Os Flow UI by typing in open flow to Access Sparks UI, type in OpenSparkUI. And now you're set up to write scholar code from within a sparkling shell.