Analysis Operations

Analysis operations combine raster and vector data to produce derived insights: zonal statistics, weighted overlays, change detection, spatial indexing, point sampling, and distance calculations.

Part of the geo domain.

Operations

Operation	Type	Description	Key Params
analysis_zonal_stats	raster + vector → table	Compute statistics of raster values within polygon zones.	stats (array: mean, min, max, std, median, etc.), admin_level
analysis_weighted_overlay	raster[] → raster	Multi-criteria weighted combination. Inputs normalized to 0-1, weights normalized to sum to 1.	weights (array, required), normalize_inputs (bool), output_range
raster_change_detection	raster + raster → raster	Pixel-wise change between baseline and current rasters. Supports difference, normalized difference, and threshold classification.	method, threshold, min_patch_pixels
analysis_point_sample	raster + vector → table	Sample raster values at point locations.	property_name
analysis_h3_index	raster → table	Index data onto H3 hexagonal grid. Resolution auto-detected from pixel size if not specified.	resolution (0-15), stats
analysis_histogram	raster → stats	Compute histogram of raster value distribution.	bins (int)
analysis_distance_to_features	raster + vector → raster	Euclidean distance from each pixel to nearest feature.	(none)
analysis_summary	raster → stats	Summary statistics for entire raster.	(none)

Examples

Zonal Statistics by Watershed

layers:
  results/slope-by-watershed:
    type: vector
    compute:
      op: analysis_zonal_stats
      inputs:
        raster: { layer: terrain/slope }
        zones: { layer: source/watersheds }
      params:
        stats: [mean, min, max, std]
    style:
      type: fill
      property: mean
      palette: ylorbr
      table:
        columns:
          - { field: name, label: Watershed }
          - { field: mean, label: "Mean Slope", format: ".1f" }
          - { field: max, label: "Max Slope", format: ".1f" }

Weighted Multi-Criteria Overlay

layers:
  suitability/composite:
    type: raster
    compute:
      op: analysis_weighted_overlay
      inputs:
        rasters:
          - { layer: analysis/slope-suitability }
          - { layer: analysis/aspect-suitability }
          - { layer: analysis/land-suitability }
          - { layer: analysis/grid-proximity }
      params:
        weights: [0.30, 0.20, 0.25, 0.25]
        normalize_inputs: true
    style:
      palette: suitability
      colorRamp:
        - { value: 0.0, color: "#d73027" }
        - { value: 0.5, color: "#fee08b" }
        - { value: 1.0, color: "#1a9850" }
      form:
        fields:
          - name: slope_weight
            type: slider
            min: 0
            max: 1
            step: 0.05
            default: 0.30
            label: Slope Weight

Change Detection

layers:
  analysis/ndvi-change:
    type: raster
    compute:
      op: raster_change_detection
      inputs:
        baseline: { layer: analysis/ndvi-2020 }
        current: { layer: analysis/ndvi-2024 }
      params:
        method: threshold_classification
        threshold: 0.1
        min_patch_pixels: 50
    style:
      palette: rdylgn
      rescale: "-0.5,0.5"

H3 Hexagonal Aggregation

layers:
  results/site-hexes:
    type: vector
    compute:
      op: analysis_h3_index
      inputs:
        raster: { layer: suitability/composite }
      params:
        resolution: 7
        stats: [mean, min, max]
    style:
      fill-opacity: 0.6
      table:
        columns:
          - { field: h3_cell_id, label: "Hex Cell" }
          - { field: mean, label: "Mean Score", format: ".2f" }

Administrative Boundaries

When analysis_zonal_stats is used without explicit zones, users MAY specify an admin_level parameter to use platform-provided administrative boundaries:

params:
  admin_level: county  # country | state | county

The platform resolves boundaries from folia://boundaries/admin/{level} (Natural Earth / Census Bureau).