to_numpy()

This operation converts the current EmbeddingList instance into a two-dimensional (2D) NumPy array containing all vector embeddings.

Request Syntax

to_numpy()

RETURN TYPE:

np.ndarray

RETURNS:

A 2D NumPy array containing all vector embeddings in the shape (num_embeddings, dim).

EXCEPTIONS:

• ValueError:

  This exception will be raised if the current EmbeddingList instance is empty.

Examples

from pymilvus import EmbeddingList

# create an empty embedding list
embeddingList = EmbeddingList()

# add multiple vector embeddings in a batch
embeddingList.add_batch(
    embeddings=[[0.1, 0.2, 0.3, 0.4, 0.5], [0.5, 0.4, 0.3, 0.2, 0.1]]
)

embeddingList.to_numpy()