Installation (tutorial)

From visone manual
Revision as of 09:14, 9 September 2011 by Indlekof (talk | contribs) (Rserve installation)
Jump to navigation Jump to search

This trail explains how to download, install, and run visone and its various extensions. To use the core visone program you'll need only to read the first section Installing visone and, potentially, the second section on the installation of the java runtime environment. The other sections explain how to install and run several extension packages that can be used in connection with visone.


Installing visone

Installing visone is very simple. You can start it either from you local hard disk or use visone webstart. To download visone and start it from your local hard disk do the following two steps:

  1. Go to the visone download page and save the latest current version (e.g., visone.2.6.3.jar) on your hard disk, for instance, on your desktop.
  2. Execute the downloaded file (for instance, by double-clicking on it). The visone window opens and visone is ready for use!

(For larger projects, you might want to allocate more memory for the JVM [1] by executing the downloaded file from a command prompt: "java -Xmx256m -jar visone-2.6.3.jar" ...256m increases the default java heap size (128 MB) to 256 MB.)

Even simpler is starting the program via visone webstart. To do this go to the visone Website [2] and click on the link webstart visone in the green download box on the left. Tell your browser to execute the file visone.jnlp. (note: visone webstart uses a maximum heap size of 256 MB by default)

Troubleshooting (I): installing the java runtime environment (JRE)

If the visone window does not open and if you get a message similar to "cannot open file: visone-2.6.3.jar", then it is most likely that the Java Runtime Environment (JRE) is not installed on your computer. In this case do the following steps.

  1. Go to the Oracle Webpage [3].
  2. Search in the list for Java SE Runtime Environment (JRE) (version 6 or higher) and click on download. Follow the instructions.
  3. Once you have installed the JRE, do as described in the previous section.

Installing the R connection

Rserve installation

We assume that you have an R installation running (install the latest version fitting to your system from the link). For a connection between visone and R we additionally need the RServe packages. You can install this package directly from R:

  • start an R console (the R GUI under windows, or type "R" on a terminal in Linux/MacOS)
  • in the R console run the command: install.packages("Rserve", dependencies=T)
  • make sure the installation completed successfully
  • you can close R (i.e. by running the command q(), no need to save anything)
  • on Windows, it is mostly necessary to copy file Rserve.exe from library folder Rserve to the folder that contains file R.dll. If you were using the standard settings during your R installation, you should find file Rserve.exe in

C:\Users\'yourName'\Documents\Rnwin-library\2.13\Rserve and file R.dll in C:\Program Files\R\R-2.13.1\bin\i386

Now, your R should be ready for visone.

The following subsections explain how to set the R connection properties in the option dialog (menu file, options).

basic connection properties

Here you tell visone where your R installation resides. On some systems this is not problematic, sometimes you have to give the exact place where R is installed. If you try this on a Linux or MacOS, try not to touch anything. If this does not work for you, configure the following options:

  • "use R in system path" - use this option if you have a regular R installation on a Linux or MacOS and the binary is in the path
  • "path to R/Rserve dir" - if the first option does not work for you (or you want to start a specific R installation) give the path to the executable here (the directory where the executable "RServe.exe" (Windows) or "R" (all other) file is located)
  • "use Rserve" - if you are on a windows system, R needs to be started using a specific command, in this case check this box and give the path to the "Rserve.exe" to the option above

Sometimes it is necessary to copy the file "Rserve.exe" from the library section (i.e. C:\Users\Peter\Documents\R) to the bin section (i.e. C:\Program Files\R\R-2.12.2\bin\i386) of your R installation.

library configuration

In some cases R has a problem to find its libraries when started from visone. If you experience problems with library installation, i.e. visone connects, but cannot install libraries or send graphs to R, you can configure the location of your libraries here. This is also an option to have a unique place where all the visone installed R libraries will reside. The option is

  • "library path" - paste the path were your libraries are installed or should be installed. Use "NULL" for default (system) path.

start your own R

If you're planning to start R on a remote machine or the other options do not work for you, you can run your own instance of R and have visone connect to it:

  • "don't start R, but use" - check this if you would like visone to use an R instance started by you
  • "server or ip" - give the name of the server or its ip adress
  • "port" - the port visone should connect to

In case you have trouble connecting, you can always start your own R-session. To do this, start the usual R console, make sure your completed the RServe installation (see above) and run:

library(Rserve)
Rserve()

Starting visone on a computer prepared in this way should give you a proper connection to R.

Learn more about R - especially RSiena and some more on the connection options in a short guide on how to use Rsiena from within visone.

KNIME installation

Get a copy of Knime from http://www.knime.org and install it.

Now we need to install the "bison-plugin", go to "Help->Install new software" add "http://tech.knime.org/update/bison" as update site and install the BisoNet plugin, especially "visone Integration" is needed.

After this installation you should find a number of new nodes in the "BisoNet"-category which can be used to create, manipulate or analyze networks. In the subcategory "visone" you find the "Input Connector"-Node which is used to receive networks from visone and the "Output-Connector"-Node, used to send networks to visone.

For a test, keep your Knime running, add an "Input Connector"-Node and start visone. See further details in the page Knime menu.