te_schemas.error_recode

Error recode schemas for SDG 15.3.1 land degradation analysis.

This module defines the data structures for error recoding polygons used to correct misclassified pixels in land degradation indicator calculations. Error recoding allows users to manually override automated classifications based on local knowledge or ground-truth data.

Classes:

ErrorRecodeProperties: Properties for a single error recode polygon. ErrorRecodeFeature: A GeoJSON Feature containing an error recode polygon. ErrorRecodePolygons: A GeoJSON FeatureCollection of error recode polygons.

class te_schemas.error_recode.ErrorRecodeFeature

A GeoJSON Feature representing a single error recode polygon.

Attributes:

geometry: GeoJSON geometry dict (typically a Polygon). properties: Error recode properties defining the recoding behavior. type: Must be “Feature” per GeoJSON spec.

class Meta
unknown = 'exclude'
__init__(geometry, properties, type)
Parameters:
Return type:

None

geometry: dict
properties: ErrorRecodeProperties
type: str
class te_schemas.error_recode.ErrorRecodePolygons

A GeoJSON FeatureCollection of error recode polygons.

Used to define areas where SDG 15.3.1 land degradation classifications should be manually corrected. Each feature specifies which pixel values to recode and for which time periods.

Attributes:

features: List of ErrorRecodeFeature objects. name: Optional name for the collection. crs: Optional coordinate reference system definition. type: Must be “FeatureCollection” per GeoJSON spec.

Class Attributes:

recode_deg_to_options: Valid values for recoding degraded pixels. recode_stable_to_options: Valid values for recoding stable pixels. recode_imp_to_options: Valid values for recoding improved pixels.

class Meta
unknown = 'exclude'
__init__(features, name, crs, type)
Parameters:
Return type:

None

crs: dict | None
features: List[ErrorRecodeFeature]
name: str | None
recode_deg_to_options: ClassVar[Tuple[int | None, ...]] = (None, -32768, 0, 1)
recode_imp_to_options: ClassVar[Tuple[int | None, ...]] = (None, -32768, -1, 0)
recode_stable_to_options: ClassVar[Tuple[int | None, ...]] = (None, -32768, -1, 1)
property recode_to_trans_code_dict: Dict[Tuple[int | None, int | None, int | None], int]

Generate a mapping from recode options to transformation codes.

Creates a dictionary that maps each unique combination of (recode_deg_to, recode_stable_to, recode_imp_to) values to a sequential integer code. Used for rasterizing error polygons.

Returns:

Dict mapping (deg_to, stable_to, imp_to) tuples to integer codes.

property trans_code_lists: Tuple[List[int], List[int], List[int], List[int]]

Generate lookup tables for recode transformations.

Creates parallel lists that map transformation codes to the target values for each pixel category. Used internally for efficient raster recoding operations.

The transformation code system uses a two-digit scheme where:
  • First digit indicates the source class:

    1 = degraded, 2 = improved, 3 = stable, 0 = unchanged

  • Second digit indicates the target class:

    1 = degraded, 2 = improved, 3 = stable, 0 = unchanged

So the codes are:

  • recode_deg_to: unchanged=10, stable=12, improved=13

  • recode_stable_to: unchanged=20, degraded=21, improved=23

  • recode_imp_to: unchanged=30, degraded=31, stable=32

Returns:
Tuple of four lists:

  • codes: Sequential integer codes (0 to N-1)

  • deg_to: Target values for degraded pixels (-9999 = no change)

  • stable_to: Target values for stable pixels (-9999 = no change)

  • imp_to: Target values for improved pixels (-9999 = no change)

Note:

-9999 is used as a sentinel value meaning “no recoding”. -32768 means “recode to nodata”.

type: str
class te_schemas.error_recode.ErrorRecodeProperties

Properties for an error recode polygon feature.

Defines what pixels should be recoded within the polygon boundary and which time periods the recoding applies to.

Attributes:

uuid: Unique identifier for the error polygon. periods_affected: List of periods to apply recoding to.

Valid values: “baseline”, “report_1”, “report_2”.

location_name: Human-readable name for the error area. area_km_sq: Area of the polygon in square kilometers. process_driving_change: Description of the process causing misclassification. basis_for_judgement: Justification for the correction. recode_deg_to: What to recode degraded pixels to.

None = no change, -32768 = no data, 0 = stable, 1 = improved.

recode_stable_to: What to recode stable pixels to.

None = no change, -32768 = no data, -1 = degraded, 1 = improved.

recode_imp_to: What to recode improved pixels to.

None = no change, -32768 = no data, -1 = degraded, 0 = stable.

stats: Optional statistics dictionary for the polygon.

class Meta
unknown = 'exclude'
__init__(uuid, periods_affected, location_name=None, area_km_sq=None, process_driving_change=None, basis_for_judgement=None, recode_deg_to=None, recode_stable_to=None, recode_imp_to=None, stats=None)
Parameters:

  • uuid (UUID)

  • periods_affected (List[str])

  • location_name (str | None)

  • area_km_sq (float | None)

  • process_driving_change (str | None)

  • basis_for_judgement (str | None)

  • recode_deg_to (int | None)

  • recode_stable_to (int | None)

  • recode_imp_to (int | None)

  • stats (dict | None)

Return type:

None

area_km_sq: float | None = None
basis_for_judgement: str | None = None
location_name: str | None = None
periods_affected: List[str]
process_driving_change: str | None = None
recode_deg_to: int | None = None
recode_imp_to: int | None = None
recode_stable_to: int | None = None
stats: dict | None = None
uuid: UUID