Getting Started

Environment Setup

To run TrainDB, the following environment is required. If you have insufficient permission to install, please ask your system administrator.

TrainDB

  • Java 11+

  • Maven 3.x

  • SQLite3

TrainDB ML Model Library

  • Python 3.x

  • SDGym

    • Using pip: pip install pomegranate==0.14.6 sdgym==0.5.0

    • Using conda: conda install -c sdv-dev -c conda-forge sdgym

  • SPFlow

Installation

Download

The public repository of TrainDB is https://github.com/traindb-project/traindb. You can download the released file from the address, or you can clone using git as shown below.

$ git clone https://github.com/traindb-project/traindb.git

Build

You can build TrainDB using Maven as shown below.

$ cd traindb
$ mvn package

Then, the file traindb-1.0.tar.gz is created in the traindb-assembly/target directory. You can unzip it as shown below.

$ tar xvfz traindb-assembly/target/traindb-1.0.tar.gz

To use ML models in TrainDB, you need to add model libraries. You can download our model library with the following command.

$ cd traindb-assembly/target/traindb-1.0
$ svn co https://github.com/traindb-project/traindb-model/trunk/models