Reclassify

Remap raster values to new classes based on a classification table. Supports manual breaks (value ranges mapped to new values), equal interval, quantile, and direct value mapping.

ID	raster_reclassify
Domain	geo
Category	raster
Type	raster-to-raster
Version	1.0.0

view yaml ◦view github ◦

Parameters

method: string · default manual · Classification method: - manual: User-defined break points and class values - equal_interval: Divide range into N equal-width bins - quantile: Divide data into N bins with equal count - value_map: Direct value-to-value mapping
breaks: number[] · Break points for manual classification. Values <= breaks[0] get class_values[0], values in (breaks[i-1], breaks[i]] get class_values[i]. Required for method=manual.
class_values: number[] · Output class values corresponding to each bin. Length must be len(breaks) + 1 for manual method, or n_classes for equal_interval and quantile methods.
n_classes: integer · default 5 · Number of classes for equal_interval and quantile methods
value_map: object · Direct value mapping for method=value_map. Keys are source values (as strings), values are target values. Example: {"1": 10, "2": 20}
labels: string[] · Optional labels for each class (for metadata/legend)
nodata_class: number · Value to assign to nodata pixels (default: keep as nodata)
crosswalk: string · Crosswalk ID to resolve from the registry. Resolved to a value_map at execution time via crosswalk_to_value_map(). Implies method=value_map. If both crosswalk and value_map are provided, value_map takes precedence. Example: "nlcd-to-esa-worldcover"
include_broad: boolean · default false · When using crosswalk, include broad-predicate mappings in addition to exact and close matches (default: only exact + close).