> ## Documentation Index
> Fetch the complete documentation index at: https://docs.wherobots.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Data Models

<Badge color="purple">Private Preview</Badge>

## DatasetEnum

Built-in imagery configurations available for RasterFlow mosaic creation.

### Values

<ResponseField name="NAIP_30CM" type="str">
  National Agriculture Imagery Program aerial imagery at 30 cm resolution, with red, green, blue, and near-infrared bands.
</ResponseField>

<ResponseField name="NAIP_60CM" type="str">
  National Agriculture Imagery Program aerial imagery at 60 cm resolution, with red, green, blue, and near-infrared bands.
</ResponseField>

<ResponseField name="S2_BEST_SCENE_WINDOWED_PIXEL" type="str">
  Sentinel-2 L2A composite that selects the closest valid observation for each pixel within the requested time window. Provides B02, B03, B04, and B08 at 10 m resolution.
</ResponseField>

<ResponseField name="S2_MED_HARVEST" type="str">
  Sentinel-2 L2A median composite for a latitude-based harvest-season window. Provides the B02, B03, B04, and B08 bands at 10 m resolution, plus an `N_VALID_PIXELS` band.
</ResponseField>

<ResponseField name="S2_MED_PLANTING" type="str">
  Sentinel-2 L2A median composite for a latitude-based planting-season window. Provides the B02, B03, B04, and B08 bands at 10 m resolution, plus an `N_VALID_PIXELS` band.
</ResponseField>

<ResponseField name="S2_MED_WINDOWED_PIXEL" type="str">
  Sentinel-2 L2A median composite within the requested time window. Provides B02, B03, B04, and B08 at 10 m resolution.
</ResponseField>

<ResponseField name="S2_MED_WINDOWED_PIXEL_ALL_BANDS" type="str">
  Sentinel-2 L2A median composite within the requested time window. Provides all 12 L2A surface-reflectance bands, resampled to 10 m resolution.
</ResponseField>

## DistanceMetricEnum

Distance metrics for comparing embedding vectors across time.

### Values

<ResponseField name="ALPHA_EARTH_COSINE" type="str">
  Dequantize and normalize AlphaEarth int8 embeddings before computing cosine distance.
</ResponseField>

<ResponseField name="COSINE" type="str">
  Compute one minus the cosine similarity of normalized embedding vectors.
</ResponseField>

<ResponseField name="EUCLIDEAN" type="str">
  Compute straight-line distance between embedding vectors.
</ResponseField>

## GeometryActorEnum

Model task types that produce georeferenced geometries from raster inputs.

### Values

<ResponseField name="TEXT_TO_BOUNDING_BOXES" type="str">
  Detect objects described by text prompts and return bounding-box geometries.
</ResponseField>

<ResponseField name="TEXT_TO_VECTOR_GEOMETRIES" type="str">
  Detect objects described by text prompts and return polygon geometries.
</ResponseField>

## GeometryModelRecipes

Preconfigured RasterFlow model recipes for text-prompted geometry outputs.

### Values

<ResponseField name="SAM3_TEXT_BBOX" type="str">
  Detect objects described by text prompts in 30 cm NAIP imagery and return georeferenced bounding boxes with confidence scores.
</ResponseField>

<ResponseField name="SAM3_TEXT_GEOMETRY" type="str">
  Detect objects described by text prompts in 30 cm NAIP imagery and return georeferenced polygon geometries with confidence scores.
</ResponseField>

## GeometryOutput

```python theme={"system"}
GeometryOutput(**data: Any)
```

Output reference for geometry inference workflow artifacts.

Attributes

## InferenceConfig

```python theme={"system"}
InferenceConfig(**data: Any)
```

Configuration for model inference in RasterFlow.

Attributes

## MergeModeEnum

Methods for assembling predictions from image patches.

### Values

<ResponseField name="CLIP" type="str">
  Remove the configured border from each patch before assembling predictions.
</ResponseField>

<ResponseField name="NONE" type="str">
  Assemble patch predictions without overlap-specific clipping or blending.
</ResponseField>

<ResponseField name="WEIGHTED_AVERAGE" type="str">
  Blend overlapping patch predictions with spatial weights to reduce seams.
</ResponseField>

## ModelRecipes

Preconfigured RasterFlow model recipes for raster prediction outputs.

### Values

<ResponseField name="CHESAPEAKE_RSC" type="str">
  Detect rural roads in 1 m NAIP imagery with ChesapeakeRSC semantic segmentation.
</ResponseField>

<ResponseField name="FTW" type="str">
  Detect agricultural fields and field boundaries with Fields of the World, using Sentinel-2 planting- and harvest-season composites.
</ResponseField>

<ResponseField name="META_CHM_V1" type="str">
  Estimate tree canopy height in meters from 60 cm NAIP RGB imagery with Meta CHM v1.
</ResponseField>

<ResponseField name="TILE_2_NET" type="str">
  Detect sidewalks, roads, and crosswalks in 30 cm NAIP RGB imagery with Tile2Net semantic segmentation.
</ResponseField>

## MosaicResult

Result for workflows that produce a mosaic index.

### Attributes

<ResponseField name="mosaic_index_uri" type="any">
  URI of the mosaic index GeoParquet output.
</ResponseField>

<ResponseField name="first_row_mosaic" type="any">
  URI of the first mosaic store referenced by the index, when available.
</ResponseField>

<ResponseField name="mosaic_index_gdf" type="any">
  Lazily loaded GeoDataFrame read from `mosaic_index_uri`. Requires GeoPandas.
</ResponseField>

## MosaicToMosaicActorEnum

Model task types that produce raster prediction outputs from raster inputs.

### Values

<ResponseField name="REGRESSION_PYTORCH" type="str">
  Produce continuous numeric predictions for each output pixel.
</ResponseField>

<ResponseField name="SEMANTIC_SEGMENTATION_CHANGE_DETECTION_PYTORCH" type="str">
  Stack multiple time steps as model inputs and produce one semantic-segmentation output for change detection.
</ResponseField>

<ResponseField name="SEMANTIC_SEGMENTATION_PYTORCH" type="str">
  Produce per-pixel class probabilities with a semantic-segmentation model.
</ResponseField>

<ResponseField name="TEMPORAL_MASKED_INPUTS_EMBEDDING_ACTOR" type="str">
  Produce embedding bands from time-stamped imagery with a temporal embedding model.
</ResponseField>

## ResamplingMethod

Methods for calculating output pixels when RasterFlow resizes imagery.

### Values

<ResponseField name="AVERAGE" type="str">
  Compute the weighted average of contributing pixels that contain data.
</ResponseField>

<ResponseField name="BILINEAR" type="str">
  Interpolate from nearby source pixels with a bilinear kernel.
</ResponseField>

<ResponseField name="CUBIC" type="str">
  Interpolate from nearby source pixels with a cubic convolution kernel.
</ResponseField>

<ResponseField name="CUBIC_SPLINE" type="str">
  Interpolate from nearby source pixels with a cubic B-spline kernel.
</ResponseField>

<ResponseField name="GAUSS" type="str">
  Apply a Gaussian kernel when downsampling imagery.
</ResponseField>

<ResponseField name="LANCZOS" type="str">
  Interpolate from nearby source pixels with a Lanczos windowed-sinc kernel.
</ResponseField>

<ResponseField name="NEAREST" type="str">
  Select the nearest source pixel. This is the fastest option and preserves categorical values.
</ResponseField>

<ResponseField name="RMS" type="str">
  Compute the root mean square of contributing pixels that contain data.
</ResponseField>

## SemSegRasterioConfig

```python theme={"system"}
SemSegRasterioConfig(**data: Any)
```

!!! abstract "Usage Documentation" [Models](../concepts/models.md)

A base class for creating Pydantic models.

Attributes: **class\_vars**: The names of the class variables defined on the model. **private\_attributes**: Metadata about the private attributes of the model. **signature**: The synthesized `__init__` \[`Signature`]\[inspect.Signature] of the model.

**pydantic\_complete**: Whether model building is completed, or if there are still undefined fields. **pydantic\_core\_schema**: The core schema of the model. **pydantic\_custom\_init**: Whether the model has a custom `__init__` function. **pydantic\_decorators**: Metadata containing the decorators defined on the model. This replaces `Model.__validators__` and `Model.__root_validators__` from Pydantic V1. **pydantic\_generic\_metadata**: Metadata for generic models; contains data used for a similar purpose to **args**, **origin**, **parameters** in typing-module generics. May eventually be replaced by these. **pydantic\_parent\_namespace**: Parent namespace of the model, used for automatic rebuilding of models. **pydantic\_post\_init**: The name of the post-init method for the model, if defined. **pydantic\_root\_model**: Whether the model is a \[`RootModel`]\[pydantic.root\_model.RootModel]. **pydantic\_serializer**: The `pydantic-core` `SchemaSerializer` used to dump instances of the model. **pydantic\_validator**: The `pydantic-core` `SchemaValidator` used to validate instances of the model.

**pydantic\_fields**: A dictionary of field names and their corresponding \[`FieldInfo`]\[pydantic.fields.FieldInfo] objects. **pydantic\_computed\_fields**: A dictionary of computed field names and their corresponding \[`ComputedFieldInfo`]\[pydantic.fields.ComputedFieldInfo] objects.

**pydantic\_extra**: A dictionary containing extra values, if \[`extra`]\[pydantic.config.ConfigDict.extra] is set to `'allow'`. **pydantic\_fields\_set**: The names of fields explicitly set during instantiation. **pydantic\_private**: Values of private attributes set on the model instance.

Create a new model by parsing and validating input data from keyword arguments.

Raises \[`ValidationError`]\[pydantic\_core.ValidationError] if the input data cannot be validated to form a valid model.

`self` is explicitly positional-only to allow `self` as a field name.

## TemporalScoreMethodEnum

Reference methods for scoring how each time step differs from a time series.

### Values

<ResponseField name="DIFFERENCE" type="str">
  Compare each time step with the immediately preceding time step.
</ResponseField>

<ResponseField name="LOO_MEAN" type="str">
  Compare each time step with the mean of all other time steps.
</ResponseField>

<ResponseField name="LOO_MEDOID" type="str">
  Compare each time step with the medoid of all other time steps.
</ResponseField>

## UriOutput

```python theme={"system"}
UriOutput(**data: Any)
```

Base output reference carrying a URI to a generated artifact.

Attributes

### Attributes

<ResponseField name="map_url: str | None" type="any">
  URL to view this artifact in the Wherobots Cloud map viewer.
</ResponseField>

## VectorizeMethodEnum

Methods for converting raster prediction bands to vector geometries.

### Values

<ResponseField name="SEMANTIC_SEGMENTATION_RASTERIO" type="str">
  Threshold each semantic-segmentation score band, group connected foreground pixels, and convert those groups to polygon geometries.
</ResponseField>

## VectorizeOutput

```python theme={"system"}
VectorizeOutput(**data: Any)
```

Output reference for vectorization workflows.

Attributes