CS109 Lecture 3

CS109 Lecture 3

Visualization Goal

Presentation

• Know facts about data
• Task: Communicate results

Exploration

• Data without hypothesis

• Task: Generate hypothesis

  The grestest value of a picture is when it forces us to notice what we never expected to see —John Tukey

Confirmation

• Hypothesis is given
• Task: Verify/ falsify hypothesis

The Process of data analysis

Steps

1. Ask an interesting question
2. Get the data
3. Explore the data
4. Model the data
5. Communicate and visualize the results

Note !

This is not step by step process, you can move back to every step

Data & Question

1. What are the data types ?
2. What are possible questions ?

Data Type (1)

• 1D (sequences) / 2D (maps) / 3D (shaped) / nD (relational)
• Temporal
• Trees (hierarchical)
• Networks (graphs)

Data Type (2)

• Tables
• Networks
• text

Data Type (3)

• Normal (Categorical)
• Ordinal (O)
• Quantitative (Q)
  
  • Two Type
    
    1. Interval (location of zero arbitrary)
    2. Ratio (zero fixed)

Semantics vs. Types

Data Semantics : The really-world meaning

Data Type : Interpretation in terms of scales of measurements

Data Dismensions

Univariate Data / Bivariate Data / Trivariate Data / Multivariate Data

Ps : Do not use 3D scatterplot ! And map the third dimension to some other visual attribute

Data Reduction

• Flitering
  
  • Eliminate some items or attributes
• Aggregation
  
  • Represent a group of elements by a new derived element

Statistical Graph Type

Choosing an Appropriate Chart

Note !

Sometimes we have to deal with overplotting by changing the value of alpha