Seminar Notes: Data Structure Space
and Visualization Structure Space (With
Basics and Examples)
1. Data Structure Space (Components of Data Organization)
🔷 What is Data Structure Space?
It refers to the different ways data can be arranged or structured for visualization purposes.
This arrangement helps users interact with and interpret data more efficiently.
🧱 Basic Structures Include:
- **Ordering**: Sorting items alphabetically, numerically, or by category.
- **Grids**: Tabular arrangement of rows and columns.
- **Groups**: Clustering related data points together.
- **Hierarchies**: Tree-like structures showing parent-child relationships.
- **Networks**: Graphs connecting nodes with edges.
Manual vs. Automatic:
- **Manual Ordering**: User adjusts order (e.g., drag-and-drop a column in Excel).
- **Automatic Ordering**: Software uses algorithms (like correlation) to reorder for better
pattern visibility.
📏 Goodness of Ordering:
How to measure if an arrangement is good?
- One method: **Correlation Coefficient** between dimensions:
ρ(X, Y) = Σ[(xi - μX)(yi - μY)] / (n - 1)σXσY
Where:
- X, Y = dimensions
- μ = mean
- σ = standard deviation
- n = number of data points
🔍 Example:
Imagine you are visualizing student marks in Math and Science:
- If you order them randomly, it's hard to see patterns.
- If you order by total marks (descending), you can easily spot top and low performers.
🔄 Search Strategies:
Finding the best order can be hard because there are many possibilities. A common method:
1. Pick two dimensions randomly.
2. Swap them and check if the new order is better.
3. Keep the change only if it improves clarity.
4. Repeat several times (heuristic approach).
📊 Visual Example:
In a heatmap:
- Ordered randomly → colors look scattered.
- Ordered by correlation → colors form smooth bands indicating clusters or trends.
2. Visualization Structure Space (Components of the Data Visualization)
🔷 What is Visualization Structure Space?
This refers to how visual elements themselves (like charts, axes, or panels) are arranged on
the screen. It affects how users explore and interact with the data.
🔧 Components:
- **Grid Layouts**: Aligning visuals in rows and columns.
- **Axis Adjustments**: Changing distance between axes for clarity.
- **Lens Techniques**: Using fisheye or distortion to zoom into parts of the data.
- **Component Reuse**: Applying same layout method across different datasets.
📐 Example:
- In a scatterplot matrix, you can increase spacing between axes to reduce clutter.
- In TableLens, adjusting grid sizes makes patterns more visible.
🧠 Why Important?
- Helps the user understand complex data.
- Reduces visual clutter.
- Improves interaction and navigation.
- Enables comparisons between data views.
Real Example:
- **Parallel Coordinates Plot**: If axes are too close, lines overlap. Adjusting spacing reveals
clearer paths.
- **Cluster Heatmap**: Grouped cells with borders show meaningful patterns in gene
expression or sales data.
✅ Summary:
Both Data Structure Space and Visualization Structure Space play a key role in making
visualizations readable and insightful. Structure in data helps in pattern detection, while
structure in visualization enhances understanding and usability.