pe.histogram.nearest_neighbor_backend package

Submodules

pe.histogram.nearest_neighbor_backend.auto module

pe.histogram.nearest_neighbor_backend.auto.search(syn_embedding, priv_embedding, num_nearest_neighbors, mode)[source]

Compute the nearest neighbors of the private embedding in the synthetic embedding using Faiss. If Faiss is not installed or an error occurs, fall back to the sklearn backend.

Parameters:

  • syn_embedding (np.ndarray) – The synthetic embedding

  • priv_embedding (np.ndarray) – The private embedding

  • num_nearest_neighbors (int) – The number of nearest neighbors to search

  • mode (str) – The distance metric to use for finding the nearest neighbors. It should be one of the following: “l2” (l2 distance), “cos_sim” (cosine similarity), “ip” (inner product, not supported by sklearn)

Raises:

ValueError – If the mode is unknown

Returns:

The distances and indices of the nearest neighbors

Return type:

tuple[np.ndarray, np.ndarray]

pe.histogram.nearest_neighbor_backend.faiss module

pe.histogram.nearest_neighbor_backend.faiss.search(syn_embedding, priv_embedding, num_nearest_neighbors, mode)[source]

Compute the nearest neighbors of the private embedding in the synthetic embedding using FAISS.

Parameters:

  • syn_embedding (np.ndarray) – The synthetic embedding

  • priv_embedding (np.ndarray) – The private embedding

  • num_nearest_neighbors (int) – The number of nearest neighbors to search

  • mode (str) – The distance metric to use for finding the nearest neighbors. It should be one of the following: “l2” (l2 distance), “cos_sim” (cosine similarity), “ip” (inner product)

Raises:

ValueError – If the mode is unknown

Returns:

The distances and indices of the nearest neighbors

Return type:

tuple[np.ndarray, np.ndarray]

pe.histogram.nearest_neighbor_backend.sklearn module

pe.histogram.nearest_neighbor_backend.sklearn.search(syn_embedding, priv_embedding, num_nearest_neighbors, mode)[source]

Compute the nearest neighbors of the private embedding in the synthetic embedding using sklearn.

Parameters:

  • syn_embedding (np.ndarray) – The synthetic embedding

  • priv_embedding (np.ndarray) – The private embedding

  • num_nearest_neighbors (int) – The number of nearest neighbors to search

  • mode (str) – The distance metric to use for finding the nearest neighbors. It should be one of the following: “l2” (l2 distance), “cos_sim” (cosine similarity)

Raises:

ValueError – If the mode is unknown

Returns:

The distances and indices of the nearest neighbors

Return type:

tuple[np.ndarray, np.ndarray]