Space-Time GAM: Simulated and Empirical Data for Multiscale Modeling

Alexis Comber provides data supporting a submission to the International Journal of Geographical Information Science (IJGIS). The dataset, 326.8 KB in size, includes simulated data and an empirical case study used to develop and test a novel multiscale space-time varying coefficient modeling approach using Generalized Additive Models (GAMs). The data was last updated on May 5, 2026, and is shared under a CC-BY-4.0 license.

Use Cases

• Benchmarking space-time GAMs against Multiscale Geographically and Temporally Weighted Regression based on the described empirical case study.
• Investigating space-time interaction specifications in GAM smooths based on the described model selection process.
• Analyzing the scale of space-time dependencies in predictor-response relationships based on the described modeling approach.

Strengths

• Data is directly linked to a peer-reviewed journal submission (IJGIS), suggesting academic rigor.
• Includes both simulated data and an empirical case study, allowing for method validation and application.
• The associated research compares model performance against a popular alternative (Multiscale Geographically and Temporally Weighted Regression).

Limitations

• Row count is unknown, which may limit suitability assessment for large-scale modeling.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset is small (326.8 KB), indicating limited scope for direct analysis.

Provenance

Source

figshare, authored by alexis comber.

Collection Method

Likely generated for a research study, including simulated data and an unspecified empirical case study.

Freshness

Last updated 2026-05-05 06:58:02; freshness should be verified.