WillowyBoat2388 / idiml

Modular machine learning library

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Idibon ML Core

To get started, download and install JDK 8 from Oracle: http://www.oracle.com/technetwork/java/javase/downloads/jdk8-downloads-2133151.html

Then, install Gradle 2.9 using the wrapper provided in this repository

idiml.git$ ./gradlew dependencies

Gradle Introduction

To see a list of all supported build tasks, run ./gradlew tasks from the command line.

Tasks can be specified for individual projects by prefixing the task with the project name, e.g. ./gradlew predict:jar to run the jar task within the predict project.

Compilation

You can build all of the projects by executing the jar task, either individually (like in the introduction above), or for all tasks: ./gradlew jar

Running in the Scala REPL

If you want to start the Scala REPL to try out some code quickly, you can launch it using: ./gradlew ml-app:scalaConsole -q

The REPL includes all of the idiml code dependencies in the class path, so this is a good way to quickly experiment with new libraries.

Unit testing

Run unit tests with the test task: ./gradlew test. Unit testing is performed with ScalaTest. HTML test results for each project are stored in the project's build/reports subdirectory.

Code Coverage

We use the gradle-scoverage tool for computing code coverage. It is a wrapper around scalac-scoverage-plugin. To run it:

  • ./gradlew reportScoverage will create the HTML report to view

Setting up Intellij

There are two ways to setup intellij:

  1. ./gradlew idea which will create intellij project files.
  2. Then in intellij import an existing project from the produced project files.

or

  1. Import an existing gradle project, and point it to the build.gradle file.

You will probably then need to tweak things to get say unit tests to work. To do that, click the top right hand icon next to the magnifying glass. It will show you the project structure. Make sure you have the following set:

  1. Under SDKs that Java 1.8 is set.
  2. Under global libraries you have the Scala SDK - 2.11.7.
  3. Under Modules that you have for predict, train and ml-app the compiler output is set to use module compile output path.
  4. You should now be able to run unit tests and the ml-app even.

About

Modular machine learning library

License:Other


Languages

Language:C++ 96.3%Language:Scala 3.0%Language:C 0.6%Language:Java 0.1%Language:Ruby 0.0%