@spec addItems(t(), Evision.Mat.maybe_mat_in()) :: t() | {:error, String.t()}

Add feature vectors to index.

Positional Arguments

• self: Evision.ANNIndex.t()

• features: Evision.Mat.

  Matrix containing the feature vectors to index. The size of the matrix is num_features x feature_dimension.

Python prototype (for reference only):

addItems(features) -> None

@spec build(Keyword.t()) :: any() | {:error, String.t()}

@spec build(t()) :: t() | {:error, String.t()}

Build the index.

Positional Arguments

• self: Evision.ANNIndex.t()

Keyword Arguments

• trees: integer().

  Number of trees in the index. If not provided, the number is determined automatically in a way that at most 2x as much memory as the features vectors take is used.

Python prototype (for reference only):

build([, trees]) -> None

@spec build(t(), [{:trees, term()}] | nil) :: t() | {:error, String.t()}

Build the index.

Positional Arguments

• self: Evision.ANNIndex.t()

Keyword Arguments

• trees: integer().

  Number of trees in the index. If not provided, the number is determined automatically in a way that at most 2x as much memory as the features vectors take is used.

Python prototype (for reference only):

build([, trees]) -> None

@spec create(Keyword.t()) :: any() | {:error, String.t()}

@spec create(integer()) :: t() | {:error, String.t()}

Creates an instance of annoy index class with given parameters

Positional Arguments

• dim: integer().

  The dimension of the feature vector.

Keyword Arguments

• distType: ANNIndex_Distance.

  Metric to calculate the distance between two feature vectors, can be DIST_EUCLIDEAN, DIST_MANHATTAN, DIST_ANGULAR, DIST_HAMMING, or DIST_DOTPRODUCT.

Return

• retval: Evision.ANNIndex.t()

Python prototype (for reference only):

create(dim[, distType]) -> retval

@spec create(integer(), [{:distType, term()}] | nil) :: t() | {:error, String.t()}

Creates an instance of annoy index class with given parameters

Positional Arguments

• dim: integer().

  The dimension of the feature vector.

Keyword Arguments

• distType: ANNIndex_Distance.

  Metric to calculate the distance between two feature vectors, can be DIST_EUCLIDEAN, DIST_MANHATTAN, DIST_ANGULAR, DIST_HAMMING, or DIST_DOTPRODUCT.

Return

• retval: Evision.ANNIndex.t()

Python prototype (for reference only):

create(dim[, distType]) -> retval

@spec getItemNumber(Keyword.t()) :: any() | {:error, String.t()}

@spec getItemNumber(t()) :: integer() | {:error, String.t()}

Return the number of feature vectors in the index.

Positional Arguments

• self: Evision.ANNIndex.t()

Return

• retval: integer()

Python prototype (for reference only):

getItemNumber() -> retval

@spec getTreeNumber(Keyword.t()) :: any() | {:error, String.t()}

@spec getTreeNumber(t()) :: integer() | {:error, String.t()}

Return the number of trees in the index.

Positional Arguments

• self: Evision.ANNIndex.t()

Return

• retval: integer()

Python prototype (for reference only):

getTreeNumber() -> retval

@spec knnSearch(t(), Evision.Mat.maybe_mat_in(), integer()) ::
  {Evision.Mat.t(), Evision.Mat.t()} | {:error, String.t()}

Performs a K-nearest neighbor search for given query vector(s) using the index.

Positional Arguments

• self: Evision.ANNIndex.t()

• query: Evision.Mat.

  The query vector(s).

• knn: integer().

  Number of nearest neighbors to search for.

Keyword Arguments

• search_k: integer().

  The maximum number of nodes to inspect, which defaults to trees x knn if not provided.

Return

• indices: Evision.Mat.t().

  Matrix that will contain the indices of the K-nearest neighbors found, optional.

• dists: Evision.Mat.t().

  Matrix that will contain the distances to the K-nearest neighbors found, optional.

Python prototype (for reference only):

knnSearch(query, knn[, indices[, dists[, search_k]]]) -> indices, dists

@spec knnSearch(
  t(),
  Evision.Mat.maybe_mat_in(),
  integer(),
  [{:search_k, term()}] | nil
) ::
  {Evision.Mat.t(), Evision.Mat.t()} | {:error, String.t()}

Performs a K-nearest neighbor search for given query vector(s) using the index.

Positional Arguments

• self: Evision.ANNIndex.t()

• query: Evision.Mat.

  The query vector(s).

• knn: integer().

  Number of nearest neighbors to search for.

Keyword Arguments

• search_k: integer().

  The maximum number of nodes to inspect, which defaults to trees x knn if not provided.

Return

• indices: Evision.Mat.t().

  Matrix that will contain the indices of the K-nearest neighbors found, optional.

• dists: Evision.Mat.t().

  Matrix that will contain the distances to the K-nearest neighbors found, optional.

Python prototype (for reference only):

knnSearch(query, knn[, indices[, dists[, search_k]]]) -> indices, dists

@spec load(t(), binary()) :: t() | {:error, String.t()}

Loads (mmaps) an index from disk.

Positional Arguments

• self: Evision.ANNIndex.t()

• filename: String.

  Filename of the index to be loaded.

Keyword Arguments

• prefault: bool.

  If prefault is set to true, it will pre-read the entire file into memory (using mmap with MAP_POPULATE). Default is false.

Python prototype (for reference only):

load(filename[, prefault]) -> None

@spec load(t(), binary(), [{:prefault, term()}] | nil) :: t() | {:error, String.t()}

Loads (mmaps) an index from disk.

Positional Arguments

• self: Evision.ANNIndex.t()

• filename: String.

  Filename of the index to be loaded.

Keyword Arguments

• prefault: bool.

  If prefault is set to true, it will pre-read the entire file into memory (using mmap with MAP_POPULATE). Default is false.

Python prototype (for reference only):

load(filename[, prefault]) -> None

@spec save(t(), binary()) :: t() | {:error, String.t()}

Save the index to disk and loads it. After saving, no more vectors can be added.

Positional Arguments

• self: Evision.ANNIndex.t()

• filename: String.

  Filename of the index to be saved.

Keyword Arguments

• prefault: bool.

  If prefault is set to true, it will pre-read the entire file into memory (using mmap with MAP_POPULATE). Default is false.

Python prototype (for reference only):

save(filename[, prefault]) -> None

@spec save(t(), binary(), [{:prefault, term()}] | nil) :: t() | {:error, String.t()}

Save the index to disk and loads it. After saving, no more vectors can be added.

Positional Arguments

• self: Evision.ANNIndex.t()

• filename: String.

  Filename of the index to be saved.

Keyword Arguments

• prefault: bool.

  If prefault is set to true, it will pre-read the entire file into memory (using mmap with MAP_POPULATE). Default is false.

Python prototype (for reference only):

save(filename[, prefault]) -> None

@spec setOnDiskBuild(t(), binary()) :: boolean() | {:error, String.t()}

Prepare to build the index in the specified file instead of RAM (execute before adding items, no need to save after build)

Positional Arguments

• self: Evision.ANNIndex.t()

• filename: String.

  Filename of the index to be built.

Return

• retval: bool

Python prototype (for reference only):

setOnDiskBuild(filename) -> retval

@spec setSeed(t(), integer()) :: t() | {:error, String.t()}

Initialize the random number generator with the given seed. Only necessary to pass this before adding the items. Will have no effect after calling build() or load().

Positional Arguments

• self: Evision.ANNIndex.t()

• seed: integer().

  The given seed of the random number generator. Its value should be within the range of uint32_t.

Python prototype (for reference only):

setSeed(seed) -> None