webGeoDS — Tools
Vector Toolkit
Checking a vector file before you trust it, and doing real analysis with it once you do. Free, no account, nothing ever leaves your browser. Want the full explanation instead of just the tool? The interactive articles walk through the same computations in Python and R.
Inspect → Validate → Check Topology → CRS → Buffer
Five tools are live — Inspect, Validate geometry, Check Topology, find/convert a CRS, and buffer points/lines/polygons at any distance.
Raster Toolkit
Reading a GeoTIFF, then computing with it. Inspect → Calculate → NDVI → Viewshed: dimensions, bands, resolution, CRS, NoData, and per-band value ranges first, then combine two bands into a computed result — a difference map, a Normalized Difference — shown as a colored overlay and downloadable as a GeoTIFF. The NDVI tool specializes that same computation for vegetation health, with red/near-infrared band presets. Viewshed answers a different question — what’s visible from a point, given the terrain in between — and runs on R (Python is shown only as reference in its companion article, since no viewshed implementation is available in this browser’s Python runtime).
Got a raster with NoData holes to fill? That’s the Raster Gap Filler, listed under GeoStatistics below, since it fills them by kriging.
GeoML Toolkit
Machine learning on spatial data — finding groups, and predicting labels. Cluster → Classify: find which points form dense groups with DBSCAN, or map a category recorded at sample points (soil type, habitat, land use) across the whole area with a Random Forest, shaded by how confident it is — both colored on the map, both downloadable. Two tools live in this family; more are planned.
Network Toolkit
Turning a line layer into a graph. Network from Lines: nodes at the line endpoints, edges along the lines, connected components colored both on the map and in a force-directed diagram — see exactly how much a snap tolerance changes how many pieces the network has. One tool live in this family; more are planned.
GeoStatistics Toolkit
Predicting a continuous surface from sparse point samples, and filling the gaps in a surface you already have. Kriging Interpolator → Raster Gap Filler: the first predicts a full surface from scattered measurements by fitting a variogram model to how similar nearby points really are, not just averaging distance- weighted neighbors; the second applies the same method to an existing raster, fitting a variogram on its own valid pixels to fill its NoData holes, optionally helped by a second, correlated band as a covariate (cokriging). Both run on R (Python is shown only as reference in their companion articles — kriging, cokriging — since neither implementation exists for this browser’s Python runtime). Two tools live in this family; more are planned.