Data Abstraction and Time-Series Data CS 4390/5390 Data Visualization Shirley Moore, Instructor...
-
Upload
mervyn-newton -
Category
Documents
-
view
213 -
download
0
Transcript of Data Abstraction and Time-Series Data CS 4390/5390 Data Visualization Shirley Moore, Instructor...
1
Data Abstractionand Time-Series Data
CS 4390/5390 Data VisualizationShirley Moore, Instructor
September 15, 2014
2
What – Why – How?
• Data abstraction is the what part of visualization design
• Why – task abstraction• How – visual encoding (e.g., marks, spatial
layout, color maps)• Goal: Understand dataset and datatype
characteristics so that we use appropriate visualization encodings and techniques
3
Dataset Types
• Munzner Chapter 2– Tables– Networks– Fields– Geometry
• Another dataset type: unstructured text
4
Tables• 2 dimensional with rows and columns
• Multidimensional
– Attribute values can also be multi-dimensional.
5
Networks and Trees
• Network
• Tree (acyclic network)
Nodes and links can both have attributes.
6
Fields• Positions with attributes• Realm of scientific visualizaiton• Types of grids:
• Considerations for continuous field data– sampling– interpolation
7
Geometry
• Positions and items
8
Attribute Data Types
For ordered data:
9
Key vs. Value Attributes
• Key– Must be uniquely valued– Can be comprised of multiple attrivbutes– Can be implicit (e.g., row number)– Also called an independent variable
• Value– Need not be uniquely valued– Can be multidimensional data
• Scalar• Vector• Tensor
– Also called a dependent variable
10
Temporal Data
• Values having to do with dates and times• Can be key or value attribute• Can have complex hierarchical structure
11
Time Series Dataset
• Common type of dataset in which time is the independent variable
• Goal in visualization is to show changes and trends over time.
• Fry Chapter 4– Example: consumption of different beverages (milk,
coffee, tea) from 1910 to 2010• Lab 2 assignment– CO2 emissions since 1950 – total, by country, per capita