Spatial Data Analysis
Spatial data formats (drag drop):
- postgresql, oracle, csv, mapinfo, kml, gml, shapefile, filedgb, geojson, cartodb
Linked Spatial & Temporal Exploration
- Explore statistical results through linked maps and charts
- Analyze spatial and temporal patterns across coordinated views
- Compare grouped vs ungrouped variables
- Examine averages across space and time
- Group observations to calculate mean and standard deviation
Multivariate Pattern Detection
- Identify relationships in multivariate space
- Use scatterplot matrices with integrated bar charts
- Map and explore patterns from non-spatial clustering algorithms:
- Principal Component Analysis (PCA)
- K-means clustering
- Hierarchical clustering
- Multidimensional Scaling (MDS)
Spatial Autocorrelation & Clustering
- Detect statistically significant spatial clusters using:
- Spatial autocorrelation
- Local Moran's I cluster map
- Local G and G* statistics
- Local join count statistics (for categorical variables)
- Apply spatial clustering algorithms:
- SKATER
- REDCAP
- Max-p regions
- K-means, K-medians, K-medoids
- Spectral clustering
Spatiotemporal Change Detection
- Test whether changes over time are spatially clustered
- Use global/local differential Moran's I
- Visualize results with LISA (Local Indicators of Spatial Association) cluster maps
Multivariate Spatial Correlation
- Assess whether multiple variables are spatially clustered
- Use spatial correlation metrics (e.g., moral correlation)
Spatial Dependence Thresholds
- Use correlograms to determine distance thresholds where spatial correlation declines
- Identify points beyond which neighboring values lose statistical dependence