Merge Embeddings (GeoAI Tools)

Summary

Merges embeddings from an input embeddings feature class into target polygons based on the area of overlap. Embeddings are aggregated and merged for each target polygon based on the input features that fall within it. For example, embeddings from ZIP code polygons or point-based locations can be merged into state-level polygon embeddings.

Usage

This tool accepts two inputs, a target polygon feature layer that defines the spatial boundaries for aggregation, and a point or polygon embedding feature class generated using the Generate Embeddings Using AI Models tool. The embedding features acts as the source embeddings to be aggregated within the target polygon geometries.

Label	Explanation	Data type
Target Features	The target polygon feature layer that provides the spatial boundaries used for aggregating embeddings from the input embeddings feature class. The geometry and attributes of these features are preserved in the output, and a new embedding field will be added based on the spatial relationship with the input embedding features.	Feature Layer
Embedding Features	The input embedding feature class generated using Generate Embeddings Using AI Models tool or provided in the compatible format. The embeddings in this feature class are aggregated based on the target features. It can have either point or polygon geometries.	Feature Layer
Output Embeddings Feature Class	The output embedding feature class with the input embeddings aggregated as per the feature geometry in the target features.	Feature Class

arcpy.geoai.MergeEmbeddings(target_features, embedding_features, out_embeddings_feature_class)

Name	Explanation	Data type
target_features	The target polygon feature layer that provides the spatial boundaries used for aggregating embeddings from the input embeddings feature class. The geometry and attributes of these features are preserved in the output, and a new embedding field will be added based on the spatial relationship with the input embedding features.	Feature Layer
embedding_features	The input embedding feature class generated using Generate Embeddings Using AI Models tool or provided in the compatible format. The embeddings in this feature class are aggregated based on the target features. It can have either point or polygon geometries.	Feature Layer
out_embeddings_feature_class	The output embedding feature class with the input embeddings aggregated as per the feature geometry in the target features.	Feature Class

Code sample

MergeEmbeddings example (stand-alone script)

The following example demonstrates how to use the MergeEmbeddings function.

# Name: MergeEmbeddings.py
# Description: Merges embeddings from an input Embeddings Feature Class into target polygons based on the area of overlap.
#
# Requirements: ArcGIS Pro Advanced license

# Import system modules
import arcpy

# Set local variables
target_features="US_counties"
embedding_features="US_zipCodes"
out_embeddings_feature_class=r"C:\project\MergeEmbeddings.gdb\MergeEmbeddings_points"

# Run Merge Embeddings
arcpy.geoai.MergeEmbeddings(
    target_features, embedding_features, out_embeddings_feature_class
    )

Environments

Output Coordinate System, Geographic Transformations, Extent, Parallel Processing Factor

Licensing information