Local Space-Time K Function

Ripley’s K function is a statistical approach computed on a set of point events distributed in n-dimensional space, and estimates the second-order property (variance) exhibited by the data. It takes into account 1) the number and 2) distance between the point events, and allows for quantifying how much the observed pattern deviates from randomness at multiple spatial scales. This is my implementation of the local K function within the space-time cube framework.
Source code:

Spatiotemporal Domain Decomposition

As soon as I enrolled in the Ph.D. program at UNC Charlotte in 2014, I started developing and implementing a recursive octree-based domain decomposition strategy for accelerating space-time analytics. I have used these concepts in many projects since. Source code:

Space-Time Kernel Density Estimation (STKDE)

STKDE is a temporal extension of the traditional 2D kernel density estimation that is popular for producing geographic heatmaps. It allows for spatiotemporal visualizations within the space-time cube framework.
Source code: 

Space-Time Inverse Distance Weighting (STIDW) Interpolation

STIDW extends the popular inverse distance weighting (IDW) interpolation with the temporal dimension. To determine the value of an attribute a given location and time, STIDW considers its spatial and temporal neighbors and applies a distance-decay relationship for weights. I use the spatiotemporal domain decomposition approach, together with k/d tree indexing to accelerate computation.
Source code:


My final project of the very first computer science class at UNCC (Illustrative Visualization), where I worked on visualizing spatiotemporal point data using a rotation axis. The angle by which each point is rotated depends on its timestamp. The higher the value, the larger the rotation.
Source code:


Visualization website for soccer player statistics of the 2014/2015 English Premier League season. It contains interactive graphs and map that lets users search for their favorite players or teams. Credit goes to my project team members Taylor Tillinger, Kongmeng Vang, John Vue, Kevin VanEmmerick and Ajay Sadhu. Keywords: Visual Analytics, Crossfilter, Soccer, Interactive, Query.
Source Code:


Implementation and application of the ARED algorithm to find action rules within the the Fragile States Index dataset. Credit goes to fellow team members Omar Eltayeby, Dongdong Li and Brianna Chen.
Source code:

Geovisualization in R

As teaching assistant of Dr. Eric Delmelle’s Geovisualization class, I created three exercises using R. I used a variety of resources which are properly credited in the documentation. Source codes: