Visualization Overview

From CDS 130

Jump to: navigation, search


  1. Objectives
  2. Motivation
  3. Priming questions
  4. Notes
    1. Visualization Topics
    2. Example 1
    3. Example 2
    4. Introduction
    5. Visualization Rules
    6. Types of Scientific Visualizations
    7. Computing and Visualization
    8. Pattern Recognition
    9. Pattern Recognition cont.
  5. Questions
    1. Spreadsheet visualizations
    2. Wireframe Motivation

1. Objectives

  • To explain why scientific visualization is an important aspect of computing for scientists.
  • To show examples of some of the most important types of scientific visualizations.

2. Motivation

  • Computer simulations produce lots of numbers. Visualization is one way of turning these numbers into scientific understanding.

3. Priming questions

  • What is the most memorable visualization you have ever seen? What aspect of the visualization made it memorable?
  • A 50-minute seminar on visualizations [1]. TED videos: [2] [3] [4] [5].

4. Notes

4.1. Visualization Topics

  • In this class we will primarily work with images
    • Using images to visualize numbers
    • Extracting numbers out of images (to do, for example, pattern recognition)

4.2. Example 1

After watching this video, turn away and write down all of the features that you remembered.

4.3. Example 2

After watching this video, turn away and write down all of the features that you remembered.

4.4. Introduction

  • Definition of Scientific Visualization (Brodlie, 1992 [6]; See also [7]):
Scientific Visualization is concerned with exploring data and information in such a way as to gain understanding and insight into the data. The goal of scientific visualization is to promote a deeper level of understanding of the data under investigation and to foster new insight into the underlying processes, relying on the humans' powerful ability to visualize. In a number of instances, the tools and techniques of visualization have been used to analyze and display large volumes of, often time-varying, multidimensional data in such a way as to allow the user to extract significant features and results quickly and easily.
  • Definition of Scientific Visualization (Friendly, 2008 [8]):
Scientific visualization is an interdisciplinary branch of science primarily concerned with the visualization of three dimensional phenomena (meteorological, medical, biological etc) where the emphasis is on realistic rendering of volumes, surfaces, illumination sources with a dynamic (time) component.
  • (Note that we will cover only the visualization of numbers and data and not sketches of objects such as insects or organs)

4.5. Visualization Rules

Grading or judging a data visualization is more like grading a speech than an essay: In a speech, far less weight is placed on the rules of grammar - communicating the idea is given far more weight.

  • In human language, there are many ways of communicating the same information
    • Different languages types
    • Different style (passive voice/active voice, first person/third person, etc.)
  • In visualization, there are many ways of communicating the same information
    • Different visualization types (1-D, 2-D, 3-D, animations, etc.)
    • Different style (color/greyscale, scatter plot/connected points, etc.)
  • Learning how to visualize data is much like learning a spoken language via "immersion". In this course we will cover issues related to how to create scientific visualizations. We will not cover how best to create scientific visualizations; that is, we'll cover how to hold the paintbrush and apply paint to a canvas - the rest could be covered in another course.

4.6. Types of Scientific Visualizations

Many examples here [9] and [10]

General Categories

  • 1-D (time + some quantity) - Generally each data point has one or more characteristics associated with it, e.g., height versus time.
  • 2-D - Generally each data point has two or more characteristics associated with it, e.g., height and weight
    • Vector [13]
    • Contour
    • Spectrogram
    • Spatial Image
    • Scatter plot (where neither variable is time)
  • 3-D - Generally each data point has three or more characteristics associated with it.
    • Wireframe [14] versus [15] - why would you use a wireframe instead of the full rendered image?
    • Volume [16]
  • Animations
    • Usually sequences of 2-D and 3-D visualizations

4.7. Computing and Visualization

  • We most often think of visualization as involving a computer rendering data to create an image.
  • What about a computer analyzing an image?

4.8. Pattern Recognition

  • Previously we translated science models into computational models (which involve computer algorithms)
  • How do we translate science data into scientifically useful information using computer algorithms?
  • Imagine the different colors on this image are made of different textures (different styles of carpet, for example).
  • How would determine where the sunspots were using only a sense of touch?
From on June 25 2017 12:13:17.

4.9. Pattern Recognition cont.

  • Example of basic numeric pattern recognition using a spreadsheet [17].
  • How (what algorithm) would you detect a pattern using only your eyes?


5. Questions

5.1. Spreadsheet visualizations

Which of the categories of visualizations described above are available in Excel? Are their names different?

5.2. Wireframe Motivation

  • Why would you want to use a wireframe instead of a full rendered image?
  • When would you want to use a full rendered image instead of a wireframe?
Personal tools