Software & Code
I develop statistical software and reproducible research code accompanying my methodological work in spatial statistics, high-dimensional statistical learning, spatial point processes, variational inference, and spatial data science.
GitHub
The following repositories provide code and computational material related to selected research projects.
Reproducible code for multiresolution variable selection in spatial point-process intensity models, motivated by crime and traffic-related spatial event data in St. Louis.
The project investigates spatially heterogeneous predictor effects using localized basis representations and penalized variable-selection methods.
View Repository →Research code for localized variable and neighborhood selection in spatial firearm-fatality data from the southeastern United States.
The project develops spatially adaptive variable-selection methodology for identifying geographically varying associations and local neighborhood effects.
View Repository →Code associated with variational restricted maximum likelihood estimation for Gaussian spatial models with intrinsic conditional autoregressive structure.
The repository contains methodology and computational implementation related to scalable variational approximation and spatial random-effects estimation.
View Repository →Research Connections
| Repository | Research Area | Related Work |
|---|---|---|
Multi-Resolution-Variable-Selection-Spatial-Point-Process
|
Spatial point processes, multiresolution variable selection | Multi-Resolution Analysis of Variable Selection in Spatial Intensity Function for Risky Road in St. Louis |
Local-Variable-and-Neighborhood-Selection
|
Local spatial variable selection, spatial firearm data | Local Variable and Neighborhood Selection in Spatial Firearm Fatality in Southeastern USA |
Variational-REML-Spatial
|
Variational inference, spatial mixed models | Variational Approximated Restricted Maximum Likelihood Estimation for Spatial Data |
Research code is maintained in separate repositories so that methodological implementations, simulations, data-processing workflows, and reproducibility material can evolve independently of this academic website.