See: Description
| Package | Description |
|---|---|
| ai.h2o.mojos.runtime |
Entry point package to load MOJO.
|
| ai.h2o.mojos.runtime.frame |
The package exposes classes to support frame and row creation.
|
| ai.h2o.mojos.runtime.lic |
Contains license verification related exceptions.
|
| ai.h2o.mojos.runtime.readers |
Provides multiple implementation of
MojoReaderBackend to support loading of MOJO pipelines from different representations (e.g., ZIP-file, folder, classpath). |
This document describes how to use DAI’s generated MOJO pipelines.
DAI-generated MOJO pipelines are intended to be easily embeddable in any environment including JVM. The only compilation and runtime dependency for a generated model is the MOJO runtime file (e.g., mojo2-runtime.jar) produced as the build output of this package.
This overview covers following topics:
A MOJO (Model Object, Optimized) is a binary format to represent DAI pipelines. The format support easy deployment or embedding of pipelines without need of using heavy weight DAI runtime.
Each DAI experiment result can be exported as MOJO pipeline.
The examples below describe how to create a model using DAI UI. The finished experiment shows option to build/download MOJO as a zip archive. The downloaded archive contains the following files:
| Filename | Description |
|---|---|
pipeline.mojo |
The binary representation of pipeline in MOJO format. |
mojo2-runtime.jar |
A Java library to execute exported MOJO pipeline. |
example.csv |
A generated sample of data matching training data structure. |
run_example.sh |
A Linux shell script to transform example.csv data with exported MOJO pipeline. |
README.txt |
A simple manual of using MOJO. |
Open a new terminal window and change current directory to a new experiment folder:
$ mkdir experiment && cd experiment
Create your main program in the experiment folder by creating a new file called Main.java (for example, using vim Main.java). Include the following contents.
import ai.h2o.mojos.runtime.MojoPipeline;
import ai.h2o.mojos.runtime.frame.MojoFrame;
import ai.h2o.mojos.runtime.frame.MojoFrameBuilder;
import ai.h2o.mojos.runtime.frame.MojoRowBuilder;
import ai.h2o.mojos.runtime.lic.LicenseException;
import ai.h2o.mojos.runtime.utils.CsvWritingBatchHandler;
import com.opencsv.CSVWriter;
import java.io.BufferedWriter;
import java.io.IOException;
import java.io.OutputStreamWriter;
import java.io.Writer;
public class DocsExample {
public static void main(String[] args) throws IOException, LicenseException {
// Load model and csv
final MojoPipeline model = MojoPipeline.loadFrom("pipeline.mojo");
// Get and fill the input columns
final MojoFrameBuilder frameBuilder = model.getInputFrameBuilder();
final MojoRowBuilder rowBuilder = frameBuilder.getMojoRowBuilder();
rowBuilder.setValue("AGE", "68");
rowBuilder.setValue("RACE", "2");
rowBuilder.setValue("DCAPS", "2");
rowBuilder.setValue("VOL", "0");
rowBuilder.setValue("GLEASON", "6");
frameBuilder.addRow(rowBuilder);
// Create a frame which can be transformed by MOJO pipeline
final MojoFrame iframe = frameBuilder.toMojoFrame();
// Transform input frame by MOJO pipeline
final MojoFrame oframe = model.transform(iframe);
// `MojoFrame.debug()` can be used to view the contents of a Frame
// oframe.debug();
// Output prediction as CSV
final Writer writer = new BufferedWriter(new OutputStreamWriter(System.out));
final CSVWriter csvWriter = new CSVWriter(writer);
CsvWritingBatchHandler.csvWriteFrame(csvWriter, oframe, true);
}
}
Compile the source code:
$ javac -cp mojo2-runtime.jar Main.java
Run the MOJO example:
# Linux and OS X users
$ java -cp .:mojo2-runtime.jar Main
# Windows users
$ java -cp .;mojo2-runtime.jar Main
The following output is displayed:
CAPSULE.True
0.5442205910902282