Concerns in Data Visualization

Visualization is a way of represent data and information graphically. It is used to display complex data using charts and graphs so that Data Analysis results can be better understood.

  • Recipient of visualization
  • Data size
  • Data type
  • Relations in data

Charts, Graphs and Diagrams

  • Frequency Table
  • Scatter Plot
  • Line Plot
  • Bar Chart
  • Histogram
  • Pie Chart
  • Exploded Pie Chart
  • Distribution Plot
  • Box plot
  • Stacked Bar Chart
  • Grouped Bar Chart
  • Stacked Area Chart
  • Pareto Diagram
  • Donut chart
  • Heatmap
  • Radar Chart
  • Tree-map
  • Parallel Coordinate Plot
  • Hexagonal Binning
  • Contour Plot
  • QQ-plot
  • Violin plot
  • Boxen-plot
  • Swarm plot
  • Point Plot
  • Word Cloud
  • Sunburst Chart

Types of Charts

  • Histogram: a graphical representation organizing a group of data points in continuous variables.
  • Box Plots: can be used detect Outliers in the data.

Choosing a Chart

  • Comparison
    • Among items
      • two variable per item: Variable width column chart
      • one variable per item
        • many categories: Table with embedded charts
        • few categories: Bar chart
    • Over time
      • Cyclical data: Circular area chart
      • None-cyclical data: Line chart
      • many categories: Vertical Bar chart
      • few categories: Line chart
  • Composition
    • Static
      • Share of total or percentage: Pie chart
      • Accumulation or subtraction to total: Waterfall chart
      • Components of components: Stacked 100% bar chart
      • Accumulation to total, or when absolute difference matters: Tree Map
    • Dynamic(over time)
      • Few periods:
        • Relative difference matters: Stacked 100% bar chart
        • Relative and absolute difference matters: Stacked bar chart
        • Only relative differences matter: Stacked area 100% chart
        • Both relative and differences matter: Stacked area chart
      • many periods:
  • Distribution
    • Single Variable
      • Single data point: Bar Histogram
      • many data points: Line Histogram
    • Two variables: Scatter Plot
  • Relationship
    • Two variables: Scatter Plot
    • Three or more variables: Scatter Plot bubble size

data-visualisation-infographic.jpg