Density, clusters and spatial statistics

Where objects thicken and how unlikely that is by chance: clustering, density grids, kernel density estimation, hot spots, Moran's I, IDW interpolation and spatial regression. The outputs are ready-made grids and numeric metrics you can drop straight into a choropleth or a report.

Tools in this group: 7

Point clustering

cluster

Groups point features into clusters: the neighbourhood radius eps_m and the minimum cluster size min_points are set by parameters.

Input: bounding boxOutput: GeoJSONfree

Calculate right here

The map appears here once the page has loaded. Without JavaScript you can still order the calculation with the API example below.
Calculation frame, WGS84

Density grid

density_grid

Counts the features of a layer in each cell of a regular grid of a given size — the base for a density map.

Input: bounding boxOutput: grid (JSON)free

Calculate right here

The map appears here once the page has loaded. Without JavaScript you can still order the calculation with the API example below.
Calculation frame, WGS84

Heatmap (KDE)

heatmap

Builds a kernel density estimate grid over the features of a layer; the grid size is set by grid_size. The output is per-cell values in JSON.

Input: bounding boxOutput: JSON metricsfree

Calculate right here

The map appears here once the page has loaded. Without JavaScript you can still order the calculation with the API example below.
Calculation frame, WGS84

Hotspots

hotspot

Finds the cells where the concentration of features is above a threshold and returns them as polygons — the zones of clustering.

Input: bounding boxOutput: GeoJSONfree

Calculate right here

The map appears here once the page has loaded. Without JavaScript you can still order the calculation with the API example below.
Calculation frame, WGS84

IDW interpolation

interpolation

Interpolates the density of point features onto a regular grid by inverse distance weighting; the strength of influence is set by power.

Input: bounding boxOutput: JSON metricsfree

Calculate right here

The map appears here once the page has loaded. Without JavaScript you can still order the calculation with the API example below.
Calculation frame, WGS84

Moran's I index

spatial_autocorrelation

Computes the spatial autocorrelation index over a grid of cells: whether the values are clustered in space or scattered at random.

Input: bounding boxOutput: JSON metricsfree

Calculate right here

The map appears here once the page has loaded. Without JavaScript you can still order the calculation with the API example below.
Calculation frame, WGS84

Spatial regression

spatial_regression

Estimates the relationship between two layers over a grid of cells: regression coefficients and correlation. A statistical relationship is not proof of causation.

Input: bounding boxOutput: JSON metricsfree

Calculate right here

The map appears here once the page has loaded. Without JavaScript you can still order the calculation with the API example below.
Calculation frame, WGS84

Input, output and limits

The input is a rectangular bounding box in WGS84 plus one of nine data layers; some tools also take a point. The output is GeoJSON, JSON metrics or a GeoTIFF.

LayerWhat it holds
buildingsbuildings
roadsroads and streets
poipoints of interest: services, shops, tourism
waterwater bodies
green_areasgreen areas: parks, forests, lawns
railwaysrailways
boundariesadministrative boundaries
landuseland use
waterwayswaterways

Every tool in the section is free. The bounding box may cover up to 100 km². Rate limits are 10 requests per minute per tool run and 30 requests per minute for reading the catalogue. Without signing in you get up to 10 successful calculations a day; after that the service asks you to register and answers 402 with the code login_required. A box larger than the allowed area is also 402, and an unknown tool id is 404.