Return to page

Using H2O in R

A quick demo on getting H2O started in R Powered by the open-source machine learning software H2O.ai.

 

 

Talking Points:

 

Speakers:

Video Narrator

Read the Full Transcript

 

 

Running H2O Through Your R Console 

 

Video Narrator

Welcome to H2O and R. The first thing that you need to run H2O through your R console, is a file downloaded from the OX Data website. You can find it by going to 0xdata.com/h2o and clicking on the download H2O button. Once your download is completed, we need to run an instance of H2O. In your terminal window, change the working directory to the downloaded file, and start an instance of H2O by entering a simple job command. This output tells us that we've successfully started an instance of H2O. We can see the IP and port number that the instance is running on by looking at the output. Now that we have an instance of H2O running, we need to go to the R console and install H2O in R. The R package for H2O is also included in the downloaded file. I've specified an absolute path from R to the H2O R package. 

 

Establishing Communication Between R and H2O 

 

 Next, make sure that the R package is running. Now we install the dependent packages, and lastly, we establish a means of communication between the R console and the running instance of H2O. 

Note that I've asked H2O to talk to R, and now R is asking me if I'd like to install the R package that matches the version of H2O that I'm running.