Class EmbeddingConverter

java.lang.Object
com.google.adk.models.springai.EmbeddingConverter
public class EmbeddingConverter extends Object
Utility class for converting between embedding formats and performing vector operations.

This class provides helper methods for working with embeddings generated by Spring AI models, including format conversions and similarity calculations.

  • Method Summary

    Modifier and Type
    Method
    Description
    static List<Double>
    calculateSimilarities(float[] query, List<float[]> candidates)
    Calculate similarity scores between a query and all candidates.
    static double
    cosineSimilarity(float[] embedding1, float[] embedding2)
    Calculate cosine similarity between two embedding vectors.
    static org.springframework.ai.embedding.EmbeddingRequest
    createRequest(String text)
    Create an EmbeddingRequest for a single text input.
    static org.springframework.ai.embedding.EmbeddingRequest
    createRequest(List<String> texts)
    Create an EmbeddingRequest for multiple text inputs.
    static double
    euclideanDistance(float[] embedding1, float[] embedding2)
    Calculate Euclidean distance between two embedding vectors.
    static List<float[]>
    extractEmbeddings(org.springframework.ai.embedding.EmbeddingResponse response)
    Extract embedding vectors from an EmbeddingResponse.
    static float[]
    extractFirstEmbedding(org.springframework.ai.embedding.EmbeddingResponse response)
    Extract the first embedding vector from an EmbeddingResponse.
    static int
    findMostSimilar(float[] query, List<float[]> candidates)
    Find the most similar embedding from a list of candidates.
    static float[]
    normalize(float[] embedding)
    Normalize an embedding vector to unit length.
    static double[]
    toDoubleArray(float[] floatArray)
    Convert float array to double array.
    static float[]
    toFloatArray(double[] doubleArray)
    Convert double array to float array.

    Methods inherited from class Object

    clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

  • Method Details

    • createRequest

      public static org.springframework.ai.embedding.EmbeddingRequest createRequest(String text)
      Create an EmbeddingRequest for a single text input.
      Parameters:
      text - The text to embed
      Returns:
      EmbeddingRequest for the text

    • createRequest

      public static org.springframework.ai.embedding.EmbeddingRequest createRequest(List<String> texts)
      Create an EmbeddingRequest for multiple text inputs.
      Parameters:
      texts - The texts to embed
      Returns:
      EmbeddingRequest for the texts

    • extractEmbeddings

      public static List<float[]> extractEmbeddings(org.springframework.ai.embedding.EmbeddingResponse response)
      Extract embedding vectors from an EmbeddingResponse.
      Parameters:
      response - The embedding response
      Returns:
      List of embedding vectors as float arrays

    • extractFirstEmbedding

      public static float[] extractFirstEmbedding(org.springframework.ai.embedding.EmbeddingResponse response)
      Extract the first embedding vector from an EmbeddingResponse.
      Parameters:
      response - The embedding response
      Returns:
      The first embedding vector, or null if no embeddings

    • cosineSimilarity

      public static double cosineSimilarity(float[] embedding1, float[] embedding2)
      Calculate cosine similarity between two embedding vectors.
      Parameters:
      embedding1 - First embedding vector
      embedding2 - Second embedding vector
      Returns:
      Cosine similarity score between -1 and 1

    • euclideanDistance

      public static double euclideanDistance(float[] embedding1, float[] embedding2)
      Calculate Euclidean distance between two embedding vectors.
      Parameters:
      embedding1 - First embedding vector
      embedding2 - Second embedding vector
      Returns:
      Euclidean distance

    • normalize

      public static float[] normalize(float[] embedding)
      Normalize an embedding vector to unit length.
      Parameters:
      embedding - The embedding vector to normalize
      Returns:
      Normalized embedding vector

    • findMostSimilar

      public static int findMostSimilar(float[] query, List<float[]> candidates)
      Find the most similar embedding from a list of candidates.
      Parameters:
      query - The query embedding
      candidates - List of candidate embeddings
      Returns:
      Index of the most similar embedding, or -1 if no candidates

    • calculateSimilarities

      public static List<Double> calculateSimilarities(float[] query, List<float[]> candidates)
      Calculate similarity scores between a query and all candidates.
      Parameters:
      query - The query embedding
      candidates - List of candidate embeddings
      Returns:
      List of similarity scores

    • toDoubleArray

      public static double[] toDoubleArray(float[] floatArray)
      Convert float array to double array.
      Parameters:
      floatArray - The float array
      Returns:
      Equivalent double array

    • toFloatArray

      public static float[] toFloatArray(double[] doubleArray)
      Convert double array to float array.
      Parameters:
      doubleArray - The double array
      Returns:
      Equivalent float array