R Programming for Spatial Analytics Training Course

Participants can select, read and inspect appropriate vector, raster and tabular spatial data formats.

• Vector geometries: points, lines, polygons, multipolygons and geometry collections
• Raster grids, cell resolution, extent, bands and no-data values
• Read GeoPackage, GeoJSON, Shapefile, GeoTIFF and NetCDF spatial files
• Compare feature attributes, geometry columns and raster cell values
• Use GDAL drivers to identify supported spatial data formats
• Assess file metadata, licensing terms and source provenance records

Participants can identify, transform and verify coordinate reference systems for reliable spatial analysis.

• Geographic and projected coordinate reference systems distinguished through measurement purpose
• Inspect EPSG codes, WKT definitions and axis-order metadata
• Transform sf objects with st_transform and documented target CRS
• Select projected coordinate systems for area, distance and direction calculations
• Diagnose misplaced layers caused by incorrect or missing CRS metadata
• Measure positional accuracy, scale effects and coordinate precision limits

Participants can clean, transform and combine vector datasets using sf and tidyverse methods.

• Create sf objects from coordinate columns and known coordinate systems
• Validate polygon geometries with st_is_valid and st_make_valid
• Repair duplicated features, empty geometries and invalid ring structures
• Filter, mutate and summarise spatial attributes with dplyr verbs
• Cast geometry types using st_cast for analysis-ready feature layers
• Write cleaned vector layers to GeoPackage with explicit layer names

Participants can calculate spatial relationships and produce vector-derived analytical datasets.

• Spatial predicates identify intersects, contains, touches, crosses and within relationships
• Join attributes by location using st_join and explicit predicates
• Buffer facilities using st_buffer with suitable projected coordinate systems
• Intersect boundaries and assets using st_intersection for overlay analysis
• Calculate nearest-feature distances with st_nearest_feature and st_distance
• Dissolve administrative units using group_by and summarise geometry operations

Participants can generate proximity and accessibility indicators for service, asset and population analysis.

• Calculate polygon areas and perimeters using appropriate projected coordinate systems
• Build distance bands around assets with multi-ring buffer methods
• Estimate population coverage through boundary overlays and weighted aggregation
• Compare Euclidean distance assumptions with road-network travel-time measures
• Identify underserved zones using threshold-based access indicators
• Map service catchments with transparent assumptions and sensitivity checks

Participants can prepare raster datasets for spatial analysis using terra workflows.

• Create SpatRaster objects from single-band and multiband GeoTIFF files
• Inspect raster extent, resolution, coordinate system and layer names
• Crop and mask imagery using vector areas of interest
• Resample rasters using nearest-neighbour and bilinear interpolation methods
• Align grids before cell-by-cell comparison or raster algebra
• Manage no-data values, categorical classes and raster attribute tables

Participants can derive and interpret spectral indices and classified land-cover outputs.

• Band combinations for Sentinel-2 and Landsat multispectral imagery interpretation
• Calculate NDVI, NDWI and NDBI using raster algebra
• Apply cloud masks and quality-assurance bands before index calculation
• Classify land cover using supervised training samples and random forests
• Assess classification accuracy with confusion matrices and validation samples
• Summarise land-cover composition within administrative or project boundaries

Participants can quantify raster change and summarise raster indicators within defined zones.

• Difference vegetation indices between dates using aligned raster stacks
• Detect threshold-based change areas and remove isolated raster cells
• Calculate zonal statistics for districts, catchments and protected areas
• Extract raster values at monitoring points and survey locations
• Aggregate temporal raster series into seasonal and annual indicators
• Report uncertainty arising from resolution, timing and classification error

Participants can produce clear static maps that communicate spatial patterns and uncertainty.

• Map vector layers with ggplot2, geom_sf and coordinate-aware scales
• Select sequential, diverging and categorical colour schemes for mapped variables
• Classify choropleth data using quantiles, natural breaks and fixed thresholds
• Add labels, scale bars, north arrows and source notes
• Represent uncertainty through transparency, hatching and confidence intervals
• Apply cartographic hierarchy to titles, legends and visual emphasis

Participants can create interactive maps and communicate analytical findings to non-specialist audiences.

• Build Leaflet maps with basemaps, popups, legends and layer controls
• Configure interactive feature styling from attribute values and classification rules
• Publish self-contained HTML maps with embedded data and dependencies
• Select map interactions appropriate for operational and executive audiences
• Write map annotations that distinguish observation, inference and recommendation
• Compare static reporting maps with exploratory interactive map products

Participants can assess clustering, dispersion and spatial dependence in point and area data.

• Construct point patterns with study windows and observation boundaries
• Calculate nearest-neighbour indices for clustered or dispersed event locations
• Estimate kernel density surfaces using bandwidth selection principles
• Measure global Moran's I for spatial autocorrelation in area data
• Map local indicators of spatial association using LISA statistics
• Interpret significance tests while accounting for multiple spatial comparisons

Participants can generate and assess interpolated surfaces from sampled point observations.

• Prepare sampled point data for interpolation and exploratory variogram analysis
• Apply inverse distance weighting with documented power and neighbourhood settings
• Fit empirical variograms using range, sill and nugget parameters
• Generate ordinary kriging predictions and associated prediction variance
• Compare interpolation methods using holdout validation and error metrics
• Mask prediction surfaces to valid geographic study areas

Participants can identify spatial dependence and fit appropriate models for area-based outcomes.

• Define spatial weights matrices using contiguity and distance-based neighbours
• Test regression residuals for remaining spatial autocorrelation patterns
• Fit spatial lag models for diffusion and interaction processes
• Estimate spatial error models when omitted effects are spatially correlated
• Compare ordinary least squares and spatial models using diagnostics
• Communicate coefficient interpretation, uncertainty and model limitations clearly

Participants can automate spatial workflows and implement quality checks for repeatable outputs.

• Structure analysis stages as targets pipeline dependencies and outputs
• Separate raw, interim and published spatial data directories
• Parameterise dates, boundaries and thresholds in Quarto reports
• Create automated checks for CRS, geometry validity and missing attributes
• Record source metadata, processing decisions and analytical version identifiers
• Profile slow code and reduce raster memory use strategically

Participants can frame a spatial analytical question, select suitable methods and specify defensible outputs.

• Translate operational questions into measurable spatial indicators and decision rules
• Define units of analysis, temporal windows and geographic boundaries
• Match vector, raster and statistical methods to evidence requirements
• Identify confounding, spatial bias and ecological fallacy risks
• Design validation criteria for maps, models and derived indicators
• Prepare concise findings for technical reviewers and decision-makers

Participants can build, review and present a complete reproducible spatial analysis addressing a defined business or public-service question.

• Frame a decision question, spatial extent and measurable analytical objective
• Assemble vector, raster and tabular inputs with documented provenance
• Build an R pipeline for preparation, analysis, mapping and reporting
• Apply validation checks to coordinate systems, geometry and derived outputs
• Review methods, maps and limitations against peer feedback criteria
• Present findings, recommendations and reproducibility requirements to stakeholders