Time Series Down-Sampling Using Distance Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for down-sampling time series data for display, such as reduction sampling, aggregation, mode median bucket, longest line bucket, and largest triangle dynamic, often result in loss of important data points or distortion of the data, as they either discard or group points arbitrarily without considering the display area's size and resolution.
Innovation Solution
A computer-implemented method that determines the absolute distance between data points and discards points if the distance is less than a threshold value dependent on the data value range and display area size, allowing for aggressive or conservative down-sampling based on the display area's characteristics, thereby optimizing the retention of useful information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If traditional down-sampling methods (reduction sampling, aggregation, mode median bucket, etc.) are used to reduce the number of data points, then the data volume is reduced, but important data points are lost or data distortion occurs
Solution Approach 1:
The patent changes the parameter of data point selection from arbitrary or aggregation-based to distance-based filtering. By introducing a distance threshold parameter that compares each data point's distance to its neighbors, the system dynamically determines which points to retain based on spatial distribution rather than fixed aggregation rules, thereby preserving important data points while reducing overall volume.
Solution Approach 2:
The patent applies different treatment to different data points based on their local characteristics. Each data point is evaluated individually based on its distance to neighboring points, allowing the system to retain points that are locally significant (isolated points) while discarding points that are locally redundant (points close to others), thus maintaining local data quality where it matters most.
2Device complexity
If traditional down-sampling methods are used, then processing complexity is reduced, but data accuracy and representation quality deteriorate
Solution Approach 1:
The patent implements a self-service mechanism where each data point evaluates its own significance by calculating its distance to neighboring points. This self-evaluation approach eliminates the need for complex external aggregation logic or multiple processing passes, achieving both simplicity in implementation and high accuracy in data representation simultaneously.
3Loss of information
If all data points are retained for display, then data completeness is maintained, but display performance and rendering speed deteriorate
Solution Approach 1:
The patent extracts only the essential data points needed for accurate representation by removing redundant points that fall below the distance threshold. This extraction process maintains data completeness for visually significant points while eliminating unnecessary data that would slow down rendering, thus improving display performance without sacrificing important information.
Data Source
AI summary
A computer-implemented method of down-sampling time series data for display is described. The method comprises a step of determining an absolute distance between a previous data point in the series and the present data point, and discarding the present data point if the determined absolute distance is less than a threshold value. The threshold value is dependent on a data value range for the time series data, and a size of a display area within which the time series data is to be displayed.