1 of 22

DATA VISUALIZATION

By

V Sowjanya

Sr. Assistant Professor

Dept of Artificial Intelligence & Data Science

LAKIREDDY BALI REDDY COLLEGE OF ENGINEERING

2 of 22

At the end of this unit, Student will be able to:

  • CO3 Apply structured approach to create effective visualization. (Apply-L3)

V Sowjanya

19.8.2025

Unit-III Outcomes

LBRCE

DATA VISUALIZATION

3 of 22

  • In this chapter an overview of a number of data and information visualization systems and toolkits is provided.
  • Concentration primarily on software that is freely available, to enable students interested in exploring further in the field of visualization to try out existing technology.

V Sowjanya

19.8.2025

Chapter Preface

LBRCE

DATA VISUALIZATION

4 of 22

V Sowjanya

19.8.2025

Types of Systems

  • Systems based on data types
  • Systems based on analysis type
  • Text Analysis and Visualization
  • Toolkits

LBRCE

DATA VISUALIZATION

5 of 22

V Sowjanya

19.8.2025

Systems Based on Data Type

  • Scientific Data
  • Multivariate Data
  • Graph Data

LBRCE

DATA VISUALIZATION

6 of 22

V Sowjanya

19.8.2025

Scientific Data

  • OpenDX which was formerly marketed as IBM Visualization Data Explorer, is an extensible visualization environment primarily used for the analysis of scientific and engineering data.
  • What separates it from most other visualization platforms is the visual programming process used to create custom visualizations.
  • Its Network Editor allows users to drag-and-drop components onto a canvas and create links between components for communicating data of specified types.
  • The modules fall into several distinct classes, including:

LBRCE

DATA VISUALIZATION

7 of 22

  • import and export—modules to load and save data in different formats;
  • flow control—modules to create loops and conditional execution;
  • realization— modules to map the data to renderable entities, such as isosurfaces, grids, and streamlines
  • rendering—modules to control display attributes, such as lighting, cameras, and clipping
  • transformation—functions to apply to the data, such as filtering, mathematical functions, and sorting
  • interactor—widgets such as file selectors, menus, dials/sliders, and button boxes.

V Sowjanya

19.8.2025

Scientific Data

LBRCE

DATA VISUALIZATION

8 of 22

V Sowjanya

19.8.2025

Scientific Data

LBRCE

DATA VISUALIZATION

9 of 22

V Sowjanya

19.8.2025

Scientific Data

LBRCE

DATA VISUALIZATION

10 of 22

  • A large range of existing modules have been developed to date, and visualizations can be created that combine different data and different mappings into a single visualization.
  • Users can either restrict themselves to using the existing library of components/modules, or they may write their own, using a software development kit that contains libraries to access data structures and interface components.

V Sowjanya

19.8.2025

Scientific Data

LBRCE

DATA VISUALIZATION

11 of 22

V Sowjanya

19.8.2025

Scientific Data

LBRCE

DATA VISUALIZATION

12 of 22

  • XmdvTool, developed at Worcester Polytechnic Institute (WPI) by Matthew Ward, Elke Rundensteiner, and their students, is a public domain visualization software package that integrates five common methods for multivariate data visualization into a single exploration application.
  • This tool includes standard scatterplot matrices, scatterplots of star glyphs, parallel coordinates, dimensional stacking, and pixel-oriented techniques.
  • These visualizations are linked together using a simple selection and highlighting mechanism called an N-dimensional brush, which defines a hyperbox in the data space.
  • Selected data in one view are also selected in other views, and the resulting selection can be highlighted, masked, deleted, or analyzed separate from the rest of the data.

V Sowjanya

19.8.2025

Multivariate Data

LBRCE

DATA VISUALIZATION

13 of 22

  • Beyond its original inception, XmdvTool has been extended to include additional architectural features for supporting large data sets.
  • First, YingHuey Fua et al. introduced hierarchical parallel coordinates for the exploration of data sets containing many records.
  • Data is hierarchically clustered and the results displayed in a summarized parallel coordinates display using variable opacity bands.

V Sowjanya

19.8.2025

Multivariate Data

LBRCE

DATA VISUALIZATION

14 of 22

V Sowjanya

19.8.2025

Multivariate Data

LBRCE

DATA VISUALIZATION

15 of 22

  • This was followed by the addition of a structure-based brush, an associated user interface for browsing, and brushing within this hierarchical data structure.
  • Jing Yang et al. generalized the application of this hierarchical data structure to XmdvTool’s other visualizations and defined the interactive hierarchical displays (IHD) framework.
  • Furthermore, XmdvTool provides a visual hierarchical dimension reduction (VHDR) framework that groups and organizes the space of dimensions, providing meaningful subspaces of the dimensions for analysis .
  • XmdvTool also includes the distance quantification classing (DQC) approach to handle nominal variables and tools to reorder dimensions to reduce visual clutter

V Sowjanya

19.8.2025

Multivariate Data

LBRCE

DATA VISUALIZATION

16 of 22

V Sowjanya

19.8.2025

Multivariate Data

LBRCE

DATA VISUALIZATION

17 of 22

  • GraphViz is a library of graph layout algorithms developed at AT&T Research.
  • The architecture and philosophy of GraphViz is rather unique compared to other visualization tools.
  • It supports a range of graph specification methods, layout methods, and rendering methods.
  • While some interactive components have been integrated with the system, it is primarily a script-driven system.
  • One selects a graph description file and feeds it to a layout engine, along with a desired output format and any other parameters to the layout technique.
  • The output formats supported are vast, to enable easy integration of results into documents, web pages, and applications.

V Sowjanya

19.8.2025

Graph Data

LBRCE

DATA VISUALIZATION

18 of 22

  • All GraphViz programs accept input files in the DOT language, which is defined by an abstract grammar. A simple example of a graph specified in DOT is as follows: digraph G { size="6,6"; node [shape=circle,fontsize=8]; rankdir=LR; st9 -> st9 [label="11/1"]; st9 -> st10 [label="10/1"]; st8 -> st8 [label="10/1"]; st8 -> st0 [label="00/-"]; st7 -> st8 [label="10/1"]; st7 -> st7 [label="00/1"]; st6 -> st6 [label="01/1"]; st6 -> st0 [label="00/-"]; st5 -> st6 [label="01/1"]; st5 -> st5 [label="11/1"]; st4 -> st4 [label="01/1"]; st4 -> st0 [label="00/-"]; st3 -> st4 [label="01/1"]; st3 -> st3 [label="00/1"]; st2 -> st9 [label="11/1"]; st2 -> st7 [label="00/1"]; st2 -> st2 [label="01/1"]; st10 -> st10 [label="10/1"]; st10 -> st0 [label="00/-"]; st1 -> st5 [label="11/1"]; st1 -> st3 [label="00/1"]; st1 -> st1 [label="10/1"]; st0 -> st2 [label="01/-"]; st0 -> st1 [label="10/-"]; st0 -> st0 [label="00/0"]; }

V Sowjanya

19.8.2025

Graph Data

LBRCE

DATA VISUALIZATION

19 of 22

  • Each line defines attributes of a graph as a whole, a node, or a link.
  • Both directed and undirected graphs are allowed (just switch -> to - for undirected).

V Sowjanya

19.8.2025

Graph Data

LBRCE

DATA VISUALIZATION

20 of 22

V Sowjanya

19.8.2025

Graph Data

LBRCE

DATA VISUALIZATION

21 of 22

  • Each of the above diagrams are generated by techniques
  • dot—a layered approach that attempts to aim edges in the same direction;
  • neato—a spring model based on multidimensional scaling;
  • circo—a circular layout that is often effective for communications networks;
  • fdp—a force-directed method that uses multigrid heuristics to enable handling of large graphs.

V Sowjanya

19.8.2025

Graph Data

LBRCE

DATA VISUALIZATION

22 of 22

  • Unit-III Outcomes
  • Chapter Preface
  • Types of Systems
  • Systems Based on Data Types
  • Scientific Data
  • Multivariate Data
  • Graph Data

V Sowjanya

19.8.2025

Summary

LBRCE

DATA VISUALIZATION