h2oai/app-mojo-servlet

Name: app-mojo-servlet

Owner: H2O.ai

Description: Example of putting a mojo zip file as a resource into a java servlet.

Created: 2017-07-21 01:24:37.0

Updated: 2017-07-21 01:30:53.0

Pushed: 2017-07-25 16:44:55.0

Homepage: null

Size: 14172

Language: Java

GitHub Committers

User	Most Recent Commit	# Commits

Other Committers

User	Email	Most Recent Commit	# Commits

README

H2O generated MOJO model REST API servlet example

This example shows a generated MOJO being called using a REST API.

Files

(Build files)

• build.gradle

(Dataset and model training script)

• data/callcenter_data.csv
• script.R

(Generated)

• src/main/resources/regression_model.zip

(Servlet Back-end)

• src/main/webapp/WEB-INF/web.xml
• src/main/java/ai/h2o/PredictRegressionServlet.java
• src/main/resources/regression_model.zip

(Output)

• build/libs/ROOT.war

Steps to run

Step 1: Install H2O's R package if you don't have it yet.

http://h2o.ai/download

Step 2: Build the model.

The model MOJO file is dropped into src/main/resources/regression_model.zip

cript script.R

Step 3: Build the java WAR file.

gradlew build

The output of this process is build/libs/ROOT.war

Step 4: Deploy the .war file in a Jetty servlet container.

This is handy for testing.

gradlew jettyRunWar

Step 5: In another terminal window, run a transaction against the jetty server to make a prediction.

 -v "http://localhost:8080/predict?priority=1&type=PS"

Data

The dataset is telephone data recorded from a call-center of ?Anonymous Bank?. The data can be found here:

• http://ie.technion.ac.il/serveng/callcenterdata/index.html