Coursework
Weekly hands-on exercises exploring geospatial data, analysis, and visualisation in R, completed for ISSS626: Geospatial Analytics and Applications at Singapore Management University.
01 Hands-on Exercise 1
Geospatial data fundamentals: importing, wrangling, and mapping spatial data with sf and tmap.
02 Hands-on Exercise 2
Spatial point patterns analysis of Singapore’s childcare centres, using kernel density estimation and distance-based tests against complete spatial randomness with spatstat.
03 Hands-on Exercise 3
Point patterns confined to a road network, using network kernel density estimation and network-constrained K-functions over Punggol with spNetwork.
ExerciseNetwork Constrained Spatial Point Patterns Analysis→
04 Hands-on Exercise 4
Areal data and neighbourhood definition: contiguity, distance-band and adaptive spatial weights over Hunan’s 88 counties, and the spatially lagged variables built from them with spdep.
05 Hands-on Exercise 5
Whether Hunan’s economic development clusters in space, and where. Global tests give one number for the province; local ones give a value per county, and the two disagree about what the pattern actually is.
06 Hands-on Exercise 6
Carving Shan State into planning regions from six ICT penetration measures. Hierarchical clustering ignores geography and fragments; SKATER enforces contiguity through a minimum spanning tree; ClustGeo trades the two off against each other. The exercise measures what that constraint costs.
ExerciseGeographical Segmentation with Spatially Constrained Clustering→
07 Hands-on Exercise 7
How far each 250 m hexagon of Singapore is from an eldercare centre, measured three ways with SpatialAcc: Hansen, KD2SFCA and SAM. With placeholder demand all three reduce to distance to the nearest centre, and five per cent of the course’s road-distance matrix turns out not to be distance.