vivaGen Files

vivaGen -- A survival data set generator


Quick Installation
==================

1. Install a Java Runtime Environment (JRE) version 8 for end users or Java Development Kit (JDK) version 8 for developers. vivagen was tested with Oracle JRE under Windows 10 and OpenJDK under CentOS Linux.

2. Download vivagen-[version].zip from https://sourceforge.net/projects/vivagen/files/ and extract it to a suitable location on your computer.

3. Start vivagen by double-clicking on vivagen.jar or start it from a command line with "java -jar vivagen.jar"

4. The graphical frontend of vivagen will open.


Detailed instructions
=====================

The file vivagen zip file contains the following files and directories:

- vivagen.jar
|-config
|  |-generateData.json
|   
|-output
   |- generatedData.csv
   |- nominalizedData.csv
 
vivagen.jar is a java archive that contains all required dependencies to run vivagen. Vivagen needs a configuration file to run. The program looks for the file in the following places:
1. You can provide the path to the config file as a command line parameter, e.g. "java -jar vivagen.jar /path/to/config/myconfig.json"
2. If no command line parameter is found, the file config/generateData.json is loaded if it exists
3. If both the files specified in 1. and 2. do not exist, a standard configuration file is loaded from the jar file.

For the initial window, the output data file specified in the configuration file is loaded. If no data file exists, a default data file is loaded from the jar file. In this case, the parameters shown in the upper frame might not match the data curves shown in the lower frame. Once the "generate and save survival data" button was pressed, both frames will be synchronized. The newly generated data files are always stored in the location specified in the configuration file. In case of the standard configuration of the jar file, output files are stored in the current directory.
Please be advised that the generation process might take several minutes (e.g. five minutes for 1000 instances) depending on the number of instances and your hardware. Most of the time is spent by the WEKA library for nominalizing the data.
Please note: You can interactively adjust parameters for the data generation process and generate data files. However, the modified settings are currently not saved to the configuration file.
 
 
Notes for developers
====================
 
For compiling vivagen, two libraries need to be included into the Java buildpath:

1. The "Weka 3: Data Mining Software in Java"-framework of at least version 3.6 is needed.
This project homepage is https://www.cs.waikato.ac.nz/ml
You can download the jar file for example from https://sourceforge.net/projects/weka/files/weka-3-9/3.9.3/
Choose the file with "zip"-extension and extract weka.jar. 

2. A GPLv2 compatible JSON implementation. One implementation is available from:
https://repo1.maven.org/maven2/org/glassfish/javax.json/1.0.4/javax.json-1.0.4.jar



The main class of vivagen is /org/cliommics/vivagen/gui/MainFrame