UNIT-I: DATA VISUALIZATION
1. Introduction to Visual Perception
Definition: Visual perception is the process of interpreting visual
information from the surrounding environment. It plays a vital role
in data visualization as it determines how effectively viewers
understand presented data.
Key Concepts:
o Human Visual System: Processes images faster than text;
~90% of information transmitted to the brain is visual.
o Preattentive Processing: Certain visual attributes are detected
almost instantly (within 250ms) without conscious effort.
Examples of Preattentive Attributes:
Color: Red vs. blue objects stand out.
Orientation: Vertical vs. horizontal lines.
Size: Larger circles attract attention.
Position: Objects at unusual positions stand out.
o Why important in Visualization?
Helps design charts that guide user attention
effectively.
Example Diagram: Show multiple shapes with one in a different
color/size to explain preattentive processing.
Python Example: Plot a scatter plot where one point is highlighted
in red to demonstrate focus.
2. Visual Representation of Data
Definition: Transforming abstract data into graphical form to make
patterns and insights visible.
Types of Data & Suitable Charts:
o Categorical: Bar charts, pie charts.
o Numerical: Histograms, line charts.
o Temporal: Time-series line charts.
o Spatial: Heatmaps, geographic maps.
Steps to Create Effective Representation:
1. Understand data type and relationship.
2. Choose the right visual encoding (position, length, angle,
area, color).
3. Reduce clutter and emphasize key points.
Example: Dataset of sales over 12 months → use a line chart, not a
pie chart.
Python Example: Line chart of monthly sales.
3. Gestalt Principles in Visualization
Gestalt Theory: Explains how humans perceive patterns and
structure in visual information.
Key Principles (with diagrams/examples):
1. Proximity: Items close to each other appear related (grouped
bar charts).
2. Similarity: Items with similar color/shape are perceived as a
group (scatter plot with colored categories).
3. Continuity: Viewers follow continuous lines/patterns (line
graphs).
4. Closure: Incomplete shapes are perceived as whole (dotted
outlines in charts still show a shape).
5. Figure-Ground: Distinguishing object (figure) from
background (ground).
Teaching Note: Ask students to identify Gestalt principles in
common dashboards.
4. Information Overload in Visualization
Definition: Presenting too much information causes confusion
rather than clarity.
Causes:
o Too many data points without aggregation.
o Overuse of colors and 3D effects.
o Poor layout (cluttered dashboards).
Solutions:
o Filtering: Show only relevant data.
o Aggregation: Summarize large datasets (e.g., monthly
average instead of daily).
o Focus+Context: Highlight key data while providing overall
context.
Example: Show a cluttered vs. clean bar chart of the same data.
5. Creating Visual Representations
Visual Encoding: Map data attributes to visual properties (position,
color, size, shape).
Best Practices:
o Use position and length (most accurate) over area or color
(less accurate).
o Avoid distortion (e.g., 3D pie charts).
Python Example: Compare two bar charts – one cluttered vs. one
optimized.
6. Visualization Reference Model
Stages:
1. Data: Raw input.
2. Transformation: Cleaning, filtering, summarizing.
3. Mapping: Assigning data to visual variables.
4. View: Generating visualization.
5. Interaction: User feedback modifies the view/data.
Diagram: A pipeline showing data → transformation → mapping →
view → interaction.
7. Visual Mapping
Definition: Linking data attributes to graphical properties.
Example: In a bubble chart:
o X-axis = GDP, Y-axis = Life Expectancy, Size = Population,
Color = Continent.
Python Example: Bubble chart with synthetic data.
8. Visual Analytics
Definition: Integration of automated data analysis (ML/statistics)
with visualization to support decision-making.
Use Case: Fraud detection dashboards, where anomalies are
highlighted automatically.
9. Designing Visualization Applications
Key Design Principles:
o Simplicity, Consistency, Interactivity.
o Align with user goals.
Common Tools: Tableau, PowerBI, Plotly, Seaborn.
Next Step:
I can now:
1. Add all these detailed sections with diagrams & Python-generated
visuals into a professionally formatted Word (.docx) for teaching.
2. Include discussion questions & student exercises at the end of the
unit.
1. INTRODUCTION TO VISUAL PERCEPTION
Visual perception is the brain’s ability to interpret and make sense of visual information
captured by the eyes.
Importance in Data Visualization:
- Helps in designing effective charts and dashboards.
- Leverages human ability to spot patterns quickly.
- Enables quick decision-making.
2. VISUAL REPRESENTATION OF DATA
Transforming raw data into a graphical format so patterns, trends, and insights are easily
visible.
Types of Data:
- Categorical: Bar charts, pie charts.
- Numerical: Histograms, scatter plots.
- Time-Series: Line charts.
- Geographical: Maps, heatmaps.
3. GESTALT PRINCIPLES IN VISUALIZATION
Gestalt principles explain how humans perceive structured patterns:
1. Proximity: Objects close together are perceived as related.
2. Similarity: Objects sharing color/shape appear grouped.
3. Continuity: We follow smooth paths/lines.
4. Closure: Incomplete shapes appear complete.
5. Figure-Ground: Distinguish object from background.
4. INFORMATION OVERLOAD
Too much data shown makes users struggle to extract insights.
Solutions:
- Filtering: Show only relevant data.
- Aggregation: Summarize datasets.
- Focus+Context: Highlight main insights with context.
5. VISUAL MAPPING
Mapping data attributes to graphical properties like position, size, color, shape.
Best Practices:
- Use position and length (accurate perception).
- Avoid misleading visuals (3D pie charts).
