The config block

The config block contains configuration options for a given visualization.

These are the fields that each config block can contain:

Field name	Description
aggregation	Mandatory for H3 visualizations. How points are aggregated within each cell: "count", "sum", "mean", "min", or "max". "sum" and "mean" also require numericAttribute.
band	Optional. Used in raster numeric visualizations. The raster band that we’ll use to get the data from (1-indexed).
baseBinLevel	Used in H3 visualizations. The H3 cell resolution (0–15). Country-scale ≈ 3–4, city-scale ≈ 6–7.
binMode	Used in H3 visualizations. "fixed" (same cell size at all zooms), "low"/"medium"/"high" (zoom-adaptive), or "auto".
categoricalAttribute	Mandatory for vector categorical visualizations. The attribute that contains the categorical values that will be used. (Raster categorical layers classify raw pixel values instead and omit this field.)
categories	Mandatory for a categorical visualization. Either an explicit array of category values, or a shortcut: {"type": "top", "count": N} for the N most common values, {"type": "bottom", "count": N} for the N least common, or {"type": "all"} (raster) for every unique pixel value.
labelAttribute	Optional. Defines which dataset attribute or attributes to use for labeling. If multiple values are provided, the first available one will be used.
method	Optional. Used in multiband raster numeric visualizations. Maps a spectral index (NDVI/NDMI/NDWI) to its band assignments. See Raster visualizations.
noData	Optional. Used in raster visualizations. A value or array of values that won’t be shown (e.g. [-9999] for voids, [0] for no-data).
numericAttribute	Mandatory for a numeric visualization. The attribute that contains the numeric values used.
otherOrder	Optional. Used in categorical visualizations. It can be set to either “below” or “above” to make features that do not match any of the defined categories render below or above the other ones. The default position is “below”.
outOfRangeValues	Optional. Used in raster numeric visualizations. "color" clamps out-of-range pixels to the nearest class color; "hide" makes them transparent.
rasterResampling	Optional. Used in raster numeric (and tinted hillshade) visualizations — not valid for categorical rasters, which always resample nearest. "nearest" preserves exact pixel values; "linear" smoothly interpolates (use for continuous imagery/elevation). Set it explicitly rather than relying on a default.
showOther	Optional. Used in categorical visualizations. If set to true it shows all features that do not match any defined category and adds an extra entry as the last item in the legend.
steps	Mandatory for a numeric visualization. Either an explicit array of break points, or a classification shortcut such as {"type": "jenks", "count": 5} or {"type": "continuous"}.

Examples

config is always an object. The examples below show the config block in isolation.

Categorical config — explicit categories

"config": {
  "labelAttribute": ["Wikipedia", "faa"],
  "categoricalAttribute": "faa",
  "categories": ["faa-code-1", "faa-code-2", "faa-code-3"],
  "showOther": true,
  "otherOrder": "above"
}

Categorical config — “top N” shortcut

"config": {
  "categoricalAttribute": "surface",
  "categories": {"type": "top", "count": 5},
  "showOther": true
}

Vector numeric config — automatic classification

"config": {
  "numericAttribute": "percentage",
  "steps": {"type": "jenks", "count": 5}
}

Raster numeric config — single band

"config": {
  "band": 1,
  "steps": {"type": "continuous"},
  "noData": [-9999],
  "rasterResampling": "nearest"
}

H3 config — aggregate points into hexes

"config": {
  "aggregation": "sum",
  "numericAttribute": "capacity_mw",
  "baseBinLevel": 4,
  "binMode": "medium",
  "steps": {"type": "quantiles", "count": 5}
}