Scatter Plot Bin Placement Algorithm for Non-Overlapping Data Visualization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional scatter plots suffer from occlusion in dense regions, where many data points overlap, hiding the true relationship between variables and limiting the viewer's ability to detect distributions, correlations, and anomalies.
Innovation Solution
A visualization technique that represents data points in bins with dynamically determined sizes based on attribute value ranges and density distribution, using a placement algorithm to position cells without overlap, allowing for non-overlapping visualization of data points and enabling easier detection of trends and correlations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If traditional scatter plots are used to display data points, then the visualization method is simple and easy to implement, but data points in dense regions overlap and occlude each other, hiding the true number of points and preventing accurate analysis
Solution Approach 1:
The visualization is segmented into multiple bins arranged in a grid structure, where each bin contains data points that fall within its spatial boundaries. This segmentation prevents overlap by assigning each data point to a specific bin, allowing the true number of points to be displayed through stacked visual representations within each bin.
Solution Approach 2:
The patent adds a third visual dimension by stacking data point representations vertically within each bin. Instead of plotting points only on a 2D plane where they can overlap, the third dimension (vertical stacking) allows multiple points to be visualized simultaneously without occlusion, enabling accurate counting and analysis of dense regions.
2Measurement precision
If data points are plotted with precise coordinates to maintain accuracy, then the positional precision is high, but points with identical or similar coordinates overlap completely, making it impossible to distinguish individual points
Solution Approach 1:
Data points are segmented into discrete bins based on their coordinate values. Points that would otherwise overlap due to identical or similar coordinates are separated into different bins or stacked within the same bin, preserving their individual identities while maintaining coordinate accuracy through bin assignment.
Solution Approach 2:
Bins serve as intermediary containers between the continuous coordinate space and the discrete visual representations. The bins mediate the placement of data points, ensuring that points with similar coordinates are properly separated or stacked rather than overlapping, thus preserving individual point distinction while maintaining coordinate precision.
3Loss of information
If all data points are displayed without occlusion to show the true distribution, then complete visibility is achieved, but the visualization complexity increases significantly
Solution Approach 1:
The visualization space is segmented into a grid of bins that automatically organize data points. This segmentation provides a structured framework that simplifies the display of all points without occlusion, as points are systematically placed in appropriate bins rather than requiring complex individual positioning algorithms.
Solution Approach 2:
By introducing vertical stacking within bins as a third dimension, the patent achieves complete visibility of all data points in a systematic and space-efficient manner. This dimensional approach allows numerous points to be displayed without occlusion while maintaining visual organization and avoiding excessive complexity.
Data Source
AI summary
To provide non-overlapping visualization of data points of a scatter plot, bins in a visualization screen are defined. The bins are defined along a first dimension by a first attribute of the data points, and along a second dimension by a second attribute of the data points. Cells representing corresponding data points in the bins are placed using a placement algorithm that positions the cells in the bins according to values of the first and second attributes of the corresponding data points. If the placement algorithm detects plural data points that have identical values of the first and second attributes, the placement algorithm groups the data points having the identical values in nearby positions surrounding a position corresponding to the identical values of the first and second attributes to avoid overlay of the data points having the identical values of the first and second attributes.