IBM Visualization Data Explorer (DX)

Data Explorer is a general-purpose software package for data visualization and analysis. It employs a data-flow driven client-server execution model and provides a visual program editor (VPE) that allows the user to create a visualization using a point and click interface. Images can be rendered using a variety of graphics accelerators, 8, 12, and 24 plane software windows or World Wide Web browsers.

1990 Census Data Home Ownership Trends Mapping Colorado
Valuing a Perpetuity Wind Veloctiy MRI of the Brain

Interacting with Images

Java Explorer provides interaction modes similar to those in Data Explorer: Rotate, Pan, Zoom, and Pick. A new interaction mode available in Java Explorer is called Orbit mode. Each of the preceeding images corresponds to a sample visualization available thru a link in this page. Click and drag inside an image to sample orbit mode.

Creating web pages from Visual programs

Starting with a visualization and assuming it's called Tech Notes:

The generated files

Data Explorer will write the following files:

The generated makefile

myprog.make contains values which you can insert using you favorite editor. A new alternative is to set app-defaults.


DXSERVER_DIR can be set to any directory as long as that directory is included in installDir/java/server/class/dxserver.paths.

Installing Java Explorer

There is no formal installation process for Java Explorer as there is with most software products including Data Explorer. These instructions assume some familiarity with UNIX and pc file systems, http servers, and browsers.

installDir refers to the location in which you've unpacked the compressed tar file. If you haven't downloaded the distribution yet, click here for a compressed tar file version, or here for a zip file.

  1. For cshell users... Add installDir/java/server/bin to your path:
    $ set path = ($path installDir/java/server/bin)
  2. For pc users... Edit startserver.bat.
  3. Have DX Installed
    You need to have a copy of Data Explorer installed somewhere on the system on which you are going to run the http daemon and the DXserver. Set DXROOT to where this version of dx is installed (e.g. /usr/lpp/dx).
  4. Install httpd (optional)
  5. Install JDK. Possible sources are JavaSoft or IBM.
  6. Install a recent version of Netscape.
    You'll need Navigator 4.04 to view all the web pages however versions as early as 3.01 may allow you to view some of the web pages. If you use Internet Explorer, you'll need 4.0 or later.

Running the Java Explorer server

You run one DXServer process on a machine and it serves all requests from all browsers pointed at the machine.

The web pages generated using Java Explorer use applets which detect the ip address of the machine from which they were loaded and attempt to contact a DXServer using that ip address.

If you are browsing web pages using a URL beginning with file:/, then the applets will be able to use only a DXServer running on the same machine as your browser.

After you've extended your path, can control the DXServer using the startserver and stopserver scripts.

Scripts for controlling DXServer

These are located in installDir/java/server/bin:

Running the Java Explorer server on a pc

A batch file is provided in installDir/java/server/bin/startserver.bat to start the pc version of DXServer. Using an editor such as vi or NotePad on your pc, edit in real values for DXROOT, JDKPATH, and JXVSPATH in this script. If you edit this file and then find that it won't run, then reinstall from the original zip file and re-edit the file using a different editor. You may have saved unprintable characters in the file.

Configuration Files

DXServer reads the following files from its current working directory (which will be installDir/java/server/class): Lines beginning with // and empty lines are ignored. DXServer rereads these files as they are updated, so it isn't necessary to restart.

Changes in this release

Changes in earlier releases.

Java Interfaces


public interface DXLinkHandler {
    public void outputHandler(String key, String msg, Object data);
    public boolean hasHandler(String key);

outputHandler() will be called once each time a value is received from Data Explorer for the corresponding DXLOutput tool. The name of the tool will be in key.

hasHandler() will be called by Java Explorer to determine if the applet intends to handle messages produced the the DXLOutput tool named in key.


Richard Brealy and Stewart Myers 1981. Principles of Corporate Finance. McGraw-Hill Book Company


You should expect to have to find workarounds for things in the 'Issues' list and software fixes in Java Explorer for things in the 'Bugs' list.



[ IBM home page | Order | Search | Contact IBM | Help | (C) | (TM) ]