ontolearn.data_struct

Data structures.

Classes

PrepareBatchOfPrediction	An abstract class representing a Dataset.
PrepareBatchOfTraining	An abstract class representing a Dataset.
Experience	A class to model experiences for Replay Memory.
TriplesData
CLIPDataset	An abstract class representing a Dataset.
CLIPDatasetInference	An abstract class representing a Dataset.
NCESBaseDataset
NCESDataset	An abstract class representing a Dataset.
NCESDatasetInference	An abstract class representing a Dataset.
ROCESDataset	An abstract class representing a Dataset.
ROCESDatasetInference	An abstract class representing a Dataset.
TriplesDataset	An abstract class representing a Dataset.

Module Contents

class ontolearn.data_struct.PrepareBatchOfPrediction(current_state: torch.FloatTensor, next_state_batch: torch.FloatTensor, p: torch.FloatTensor, n: torch.FloatTensor)

Bases: torch.utils.data.Dataset

An abstract class representing a Dataset.

All datasets that represent a map from keys to data samples should subclass it. All subclasses should overwrite __getitem__(), supporting fetching a data sample for a given key. Subclasses could also optionally overwrite __len__(), which is expected to return the size of the dataset by many Sampler implementations and the default options of DataLoader. Subclasses could also optionally implement __getitems__(), for speedup batched samples loading. This method accepts list of indices of samples of batch and returns list of samples.

Note

DataLoader by default constructs an index sampler that yields integral indices. To make it work with a map-style dataset with non-integral indices/keys, a custom sampler must be provided.

X

__len__()

__getitem__(idx)

get_all()

class ontolearn.data_struct.PrepareBatchOfTraining(current_state_batch: torch.Tensor, next_state_batch: torch.Tensor, p: torch.Tensor, n: torch.Tensor, q: torch.Tensor)

Bases: torch.utils.data.Dataset

An abstract class representing a Dataset.

All datasets that represent a map from keys to data samples should subclass it. All subclasses should overwrite __getitem__(), supporting fetching a data sample for a given key. Subclasses could also optionally overwrite __len__(), which is expected to return the size of the dataset by many Sampler implementations and the default options of DataLoader. Subclasses could also optionally implement __getitems__(), for speedup batched samples loading. This method accepts list of indices of samples of batch and returns list of samples.

Note

DataLoader by default constructs an index sampler that yields integral indices. To make it work with a map-style dataset with non-integral indices/keys, a custom sampler must be provided.

S

S_Prime

y

Negatives

X

__len__()

__getitem__(idx)

class ontolearn.data_struct.Experience(maxlen: int)

A class to model experiences for Replay Memory.

current_states

next_states

rewards

__len__()

append(e)

Append. :param e: A tuple of s_i, s_j and reward, where s_i and s_j represent refining s_i and reaching s_j.

retrieve()

clear()

class ontolearn.data_struct.TriplesData(knowledge_base_path)

Graph

triples = []

entities

relations

entity2idx

relation2idx

load_data()

static get_relations(data)

static get_entities(data)

class ontolearn.data_struct.CLIPDataset(data, embeddings, num_examples, shuffle_examples, example_sizes=None, k=5, sorted_examples=True)

Bases: torch.utils.data.Dataset

An abstract class representing a Dataset.

All datasets that represent a map from keys to data samples should subclass it. All subclasses should overwrite __getitem__(), supporting fetching a data sample for a given key. Subclasses could also optionally overwrite __len__(), which is expected to return the size of the dataset by many Sampler implementations and the default options of DataLoader. Subclasses could also optionally implement __getitems__(), for speedup batched samples loading. This method accepts list of indices of samples of batch and returns list of samples.

Note

DataLoader by default constructs an index sampler that yields integral indices. To make it work with a map-style dataset with non-integral indices/keys, a custom sampler must be provided.

data

embeddings

num_examples

shuffle_examples

example_sizes = None

k = 5

sorted_examples = True

__len__()

__getitem__(idx)

class ontolearn.data_struct.CLIPDatasetInference(data: list, embeddings, num_examples, shuffle_examples, sorted_examples=True)

Bases: torch.utils.data.Dataset

An abstract class representing a Dataset.

All datasets that represent a map from keys to data samples should subclass it. All subclasses should overwrite __getitem__(), supporting fetching a data sample for a given key. Subclasses could also optionally overwrite __len__(), which is expected to return the size of the dataset by many Sampler implementations and the default options of DataLoader. Subclasses could also optionally implement __getitems__(), for speedup batched samples loading. This method accepts list of indices of samples of batch and returns list of samples.

Note

DataLoader by default constructs an index sampler that yields integral indices. To make it work with a map-style dataset with non-integral indices/keys, a custom sampler must be provided.

data

embeddings

num_examples

shuffle_examples

sorted_examples = True

__len__()

__getitem__(idx)

class ontolearn.data_struct.NCESBaseDataset(vocab, inv_vocab, max_length)

vocab

inv_vocab

max_length

static decompose(concept_name: str) → list

Decomposes a class expression into a sequence of tokens (atoms)

get_labels(target)

class ontolearn.data_struct.NCESDataset(data, embeddings, num_examples, vocab, inv_vocab, shuffle_examples, max_length, example_sizes=None, sorted_examples=True)

Bases: NCESBaseDataset, torch.utils.data.Dataset

An abstract class representing a Dataset.

All datasets that represent a map from keys to data samples should subclass it. All subclasses should overwrite __getitem__(), supporting fetching a data sample for a given key. Subclasses could also optionally overwrite __len__(), which is expected to return the size of the dataset by many Sampler implementations and the default options of DataLoader. Subclasses could also optionally implement __getitems__(), for speedup batched samples loading. This method accepts list of indices of samples of batch and returns list of samples.

Note

DataLoader by default constructs an index sampler that yields integral indices. To make it work with a map-style dataset with non-integral indices/keys, a custom sampler must be provided.

data

embeddings

num_examples

shuffle_examples

example_sizes = None

sorted_examples = True

__len__()

__getitem__(idx)

class ontolearn.data_struct.NCESDatasetInference(data, embeddings, num_examples, vocab, inv_vocab, shuffle_examples, max_length=48, sorted_examples=True)

Bases: NCESBaseDataset, torch.utils.data.Dataset

An abstract class representing a Dataset.

All datasets that represent a map from keys to data samples should subclass it. All subclasses should overwrite __getitem__(), supporting fetching a data sample for a given key. Subclasses could also optionally overwrite __len__(), which is expected to return the size of the dataset by many Sampler implementations and the default options of DataLoader. Subclasses could also optionally implement __getitems__(), for speedup batched samples loading. This method accepts list of indices of samples of batch and returns list of samples.

Note

DataLoader by default constructs an index sampler that yields integral indices. To make it work with a map-style dataset with non-integral indices/keys, a custom sampler must be provided.

data

embeddings

num_examples

shuffle_examples

sorted_examples = True

__len__()

__getitem__(idx)

class ontolearn.data_struct.ROCESDataset(data, triples_data, num_examples, k, vocab, inv_vocab, max_length, sampling_strategy='p')

Bases: NCESBaseDataset, torch.utils.data.Dataset

An abstract class representing a Dataset.

All datasets that represent a map from keys to data samples should subclass it. All subclasses should overwrite __getitem__(), supporting fetching a data sample for a given key. Subclasses could also optionally overwrite __len__(), which is expected to return the size of the dataset by many Sampler implementations and the default options of DataLoader. Subclasses could also optionally implement __getitems__(), for speedup batched samples loading. This method accepts list of indices of samples of batch and returns list of samples.

Note

DataLoader by default constructs an index sampler that yields integral indices. To make it work with a map-style dataset with non-integral indices/keys, a custom sampler must be provided.

data

triples_data

num_examples

k

sampling_strategy = 'p'

load_embeddings(embedding_model)

set_k(k)

__len__()

__getitem__(idx)

class ontolearn.data_struct.ROCESDatasetInference(data, triples_data, num_examples, k, vocab, inv_vocab, max_length, sampling_strategy='p', num_pred_per_lp=1)

Bases: NCESBaseDataset, torch.utils.data.Dataset

An abstract class representing a Dataset.

All datasets that represent a map from keys to data samples should subclass it. All subclasses should overwrite __getitem__(), supporting fetching a data sample for a given key. Subclasses could also optionally overwrite __len__(), which is expected to return the size of the dataset by many Sampler implementations and the default options of DataLoader. Subclasses could also optionally implement __getitems__(), for speedup batched samples loading. This method accepts list of indices of samples of batch and returns list of samples.

Note

DataLoader by default constructs an index sampler that yields integral indices. To make it work with a map-style dataset with non-integral indices/keys, a custom sampler must be provided.

data

triples_data

k

sampling_strategy = 'p'

num_examples

num_pred_per_lp = 1

load_embeddings(embedding_model)

set_k(k)

__len__()

__getitem__(idx)

class ontolearn.data_struct.TriplesDataset(er_vocab, num_e)

Bases: torch.utils.data.Dataset

An abstract class representing a Dataset.

All datasets that represent a map from keys to data samples should subclass it. All subclasses should overwrite __getitem__(), supporting fetching a data sample for a given key. Subclasses could also optionally overwrite __len__(), which is expected to return the size of the dataset by many Sampler implementations and the default options of DataLoader. Subclasses could also optionally implement __getitems__(), for speedup batched samples loading. This method accepts list of indices of samples of batch and returns list of samples.

Note

DataLoader by default constructs an index sampler that yields integral indices. To make it work with a map-style dataset with non-integral indices/keys, a custom sampler must be provided.

num_e

head_idx

rel_idx

tail_idx

__len__()

__getitem__(idx)