d3m.metadata.base module

class d3m.metadata.base.DataMetadata(metadata: Dict[str, Any] = None, for_value: Any = None, *, generate_metadata: bool = True, check: bool = True, source: Any = None, timestamp: datetime.datetime = None)[source]

Bases: d3m.metadata.base.Metadata

A class for metadata for data values.

It checks all updates against container and data schemas. Note that as such empty (just created) metadata object does not validate against schemas. Consider setting required fields manually or use generate method as a helper to do so.

It has additional helper methods for operating on metadata of tabular data.

Parameters:
  • metadata (Dict[str, Any]) – Optional initial metadata for the top-level of the value.
  • for_value (Any) – Optional value to automatically generate metadata for. DEPRECATED: use explicit generate method call instead.
  • generate_metadata (bool) – Automatically generate metadata from for_value and update the metadata accordingly. DEPRECATED: use explicit generate method call instead.
  • check (bool) – DEPRECATED: argument ignored.
  • source (primitive or Any) – DEPRECATED: argument ignored.
  • timestamp (datetime) – DEPRECATED: argument ignored.
add_semantic_type(selector: Union[List[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]], Tuple[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE], ...]], semantic_type: str, *, source: Any = None, timestamp: datetime.datetime = None) → D[source]
append_columns(right: D, *, use_right_metadata: bool = False) → D[source]

Appends metadata for all columns from right to the right of this metadata.

Top-level metadata of right is ignored, not merged, except if use_right_metadata is set, in which case top-level metadata of this metadata is ignored and one from right is used instead.

check(value: Any) → None[source]

Checks that all metadata has a corresponding data in value and that every metadata value is valid according to schema. If not it raises an exception.

Parameters:value (Any) – Value to check against.
classmethod check_selector(selector: Union[List[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]], Tuple[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE], ...]], for_value: Any = None) → None[source]

Checks that a given selector is a valid selector. If selector is invalid it raises an exception.

It checks that it is a tuple or a list and currently we require that all segments of a selector are strings, integers, or a special value ALL_ELEMENTS.

Parameters:
  • selector (Tuple(str or int or ALL_ELEMENTS)) – Selector to check.
  • for_value (Any) – DEPRECATED: argument ignored.
clear(metadata: Dict[str, Any] = None, *, for_value: Any = None, generate_metadata: bool = True, check: bool = True, source: Any = None, timestamp: datetime.datetime = None) → D[source]

DEPRECATED: create a DataMetadata instance explicitly instead.

Creates and returns a new (clear) metadata object.

Parameters:
  • metadata (Dict[str, Any]) – Optional new initial metadata for the top-level of the value.
  • for_value (Any) – Optional value to automatically generate metadata for.
  • generate_metadata (bool) – Automatically generate metadata from for_value and update the metadata accordingly.
  • check (bool) – DEPRECATED: argument ignored.
  • source (primitive or Any) – DEPRECATED: argument ignored.
  • timestamp (datetime) – DEPRECATED: argument ignored.
Returns:

New metadata object.

Return type:

DataMetadata

generate(value: Any = None, *, compact: bool = False) → D[source]

Metadata about structure of data (dimensions) and structural types is generated for the value, and existing metadata is updated accordingly.

Parameters:
  • value (Any) – Value to automatically generate metadata for.
  • compact (bool) – Compact automatically generated metadata. Produces equivalent but compact metadata where equal metadata for all elements in a dimension are compacted into ALL_ELEMENTS selector segment.
Returns:

Metadata object updated with automatically generated metadata.

Return type:

DataMetadata

get_column_index_from_column_name(column_name: str, *, at: Union[List[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]], Tuple[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE], ...]] = ()) → int[source]
get_column_references_by_column_index(current_resource_id: str, *, at: Union[List[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]], Tuple[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE], ...]] = ()) → Dict[str, Dict[d3m.metadata.base.ColumnReference, List[d3m.metadata.base.ColumnReference]]][source]
get_columns_with_semantic_type(semantic_type: str, *, at: Union[List[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]], Tuple[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE], ...]] = ()) → Sequence[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]][source]
get_elements_with_semantic_type(selector: Union[List[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]], Tuple[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE], ...]], semantic_type: str) → Sequence[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]][source]
get_index_columns(*, at: Union[List[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]], Tuple[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE], ...]] = ()) → Sequence[int][source]

Returns column indices of the primary index columns.

It makes sure d3mIndex is always first listed.

has_semantic_type(selector: Union[List[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]], Tuple[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE], ...]], semantic_type: str) → bool[source]
horizontal_concat(right: D, *, use_index: bool = True, remove_second_index: bool = True, use_right_metadata: bool = False) → D[source]

Similar to append_columns, but it respects primary index columns, by default.

It is required that both inputs have the same number of samples.

insert_columns(columns: D, at_column_index: int) → D[source]

Inserts metadata for all columns from columns before at_column_index column in this metadata, pushing all existing columns to the right.

E.g., at_column_index == 0 means inserting columns at the beginning of this metadata.

Top-level metadata of columns is ignored.

list_columns_with_semantic_types(semantic_types: Sequence[str], *, at: Union[List[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]], Tuple[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE], ...]] = ()) → Sequence[int][source]

This is similar to get_columns_with_semantic_type, but it returns all column indices for a dimension instead of ALL_ELEMENTS element.

Moreover, it operates on a list of semantic types, where a column is returned if it matches any semantic type on the list.

list_columns_with_structural_types(structural_types: Union[Callable, Sequence[Union[str, type]]], *, at: Union[List[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]], Tuple[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE], ...]] = ()) → Sequence[int][source]

Returns a list of columns matching any of the structural types listed in structural_types. Matching allows subclasses of those types. structural_types can also be a function to call to check a structural type.

query_column(column_index: int, *, at: Union[List[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]], Tuple[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE], ...]] = (), ignore_all_elements: bool = False) → frozendict.FrozenOrderedDict[source]

Returns column metadata.

This assumes that column metadata is stored under (ALL_ELEMENTS, column_index), at optionally at selector, which might not necessary hold if metadata is not compacted. Consider using query_column_field.

Parameters:
  • column_index (int) – Column index to use.
  • at (Selector) – Selector at which to assume tabular metadata.
  • ignore_all_elements (bool) – By default, metadata from ALL_ELEMENTS is merged with metadata for an element itself. By setting this argument to True, this is disabled and just metadata from an element is returned.
Returns:

Metadata of a given column.

Return type:

frozendict.FrozenOrderedDict

query_column_field(column_index: int, field: str, *, at: Union[List[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]], Tuple[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE], ...]] = (), strict_all_elements: bool = True) → Any[source]

Returns field value of column metadata. Raises KeyError exception if metadata or field is not set.

field represents only top-level fields in metadata.

Parameters:
  • column_index (int) – Column index to use.
  • field (str) – A field name to query.
  • at (Selector) – Selector at which to assume tabular metadata.
  • strict_all_elements (bool) – If set, the method does not just return field value of column metadata, but checks that the value really holds for all rows matching the selector, without exception. This is helpful also if metadata is not compacted and field value is the same across all rows, but ALL_ELEMENTS metadata does not contain that field.
Returns:

A value of field of a given column.

Return type:

Any

remove(selector: Union[List[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]], Tuple[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE], ...]], *, recursive: bool = False, strict_all_elements: bool = False, for_value: Any = None, check: bool = True, source: Any = None, timestamp: datetime.datetime = None) → D[source]

Removes all metadata at selector.

Parameters:
  • selector (Tuple[Union[str, int, ALL_ELEMENTS]) – A selector to remove metadata at.
  • recursive (bool) – Should remove also all metadata under the selector?
  • strict_all_elements (bool) – If True, then when removing ALL_ELEMENTS entry, do not remove also metadata for all elements it matches.
  • for_value (Any) – DEPRECATED: argument ignored.
  • check (bool) – DEPRECATED: argument ignored.
  • source (primitive or Any) – DEPRECATED: argument ignored.
  • timestamp (datetime) – DEPRECATED: argument ignored.
Returns:

Updated metadata.

Return type:

DataMetadata

remove_column(column_index: int, *, at: Union[List[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]], Tuple[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE], ...]] = (), recursive: bool = False, strict_all_elements: bool = False, for_value: Any = None, source: Any = None, timestamp: datetime.datetime = None) → D[source]

Removes all column metadata for column column_index.

This removes column metadata under (ALL_ELEMENTS, column_index), at optionally at selector. It does not move to the left metadata for columns after the removed column. If you want that, use remove_columns.

Parameters:
  • column_index (int) – Column index to remove.
  • at (Selector) – Selector at which to assume tabular metadata.
  • recursive (bool) – Should remove also all metadata under the selector?
  • strict_all_elements (bool) – If True, then when removing ALL_ELEMENTS entry, do not remove also metadata for all elements it matches.
  • for_value (Any) – DEPRECATED: argument ignored.
  • source (primitive or Any) – DEPRECATED: argument ignored.
  • timestamp (datetime) – DEPRECATED: argument ignored.
Returns:

Updated metadata.

Return type:

Metadata

remove_columns(column_indices: Sequence[int]) → D[source]

Removes columns from metadata.

It moves to the left metadata for columns after removed columns. If you do not want that, use remove_column.

It throws an exception if no columns would be left after removing columns.

remove_semantic_type(selector: Union[List[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]], Tuple[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE], ...]], semantic_type: str, *, source: Any = None, timestamp: datetime.datetime = None) → D[source]
replace_columns(columns: D, column_indices: Sequence[int]) → D[source]

Replaces columns listed in column_indices with columns, in order, in this metadata.

column_indices and columns do not have to match in number of columns. Columns are first replaced in order for matching indices and columns. If then there are more column_indices than columns, additional column_indices columns are removed. If there are more columns than column_indices columns, then additional columns are inserted after the last replaced column.

If column_indices is empty, then the behavior is equivalent to calling append_columns.

Top-level metadata of columns is ignored.

select_columns(columns: Sequence[Union[int, str]], *, allow_empty_columns: bool = False) → D[source]

Returns a new metadata object with metadata only for given columns. Moreover, columns are renumbered based on the position in columns list. Top-level metadata stays unchanged, except for updating the length of the columns dimension to the number of columns.

So if the columns is [3, 6, 5] then output metadata will have three columns, [0, 1, 2], mapping metadata for columns 3 to 0, 6 to 1 and 5 to 2.

This allows also duplication of columns.

set_for_value(for_value: Any = None, *, generate_metadata: bool = True, check: bool = True, source: Any = None, timestamp: datetime.datetime = None) → D[source]

DEPRECATED: use generate method instead.

If generate_metadata is set, generate metadata from for_value and update the metadata accordingly.

Parameters:
  • for_value (Any) – Value to automatically generate metadata for.
  • generate_metadata (bool) – Automatically generate metadata from for_value and update the metadata accordingly.
  • check (bool) – DEPRECATED: argument ignored.
  • source (primitive or Any) – DEPRECATED: argument ignored.
  • timestamp (datetime) – DEPRECATED: argument ignored.
Returns:

Metadata object updated with automatically generated metadata.

Return type:

DataMetadata

set_table_metadata(*, at: Union[List[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]], Tuple[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE], ...]] = ()) → D[source]
update(selector: Union[List[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]], Tuple[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE], ...]], metadata: Dict[str, Any], *, for_value: Any = None, check: bool = True, source: Any = None, timestamp: datetime.datetime = None) → D[source]

Updates metadata with new metadata for data pointed to with selector.

If value of any field is NO_VALUE, that field is deleted.

It returns a copy of this metadata object with new metadata applied.

Parameters:
  • selector (Tuple(str or int or ALL_ELEMENTS)) – A selector pointing to data.
  • metadata (Dict) – A map of fields and values with metadata.
  • for_value (Any) – DEPRECATED: argument ignored.
  • check (bool) – DEPRECATED: argument ignored.
  • source (primitive or Any) – DEPRECATED: argument ignored.
  • timestamp (datetime) – DEPRECATED: argument ignored.
Returns:

Updated metadata.

Return type:

DataMetadata

update_column(column_index: int, metadata: Dict[str, Any], *, at: Union[List[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE]], Tuple[Union[int, str, d3m.metadata.base.ALL_ELEMENTS_TYPE], ...]] = (), source: Any = None, timestamp: datetime.datetime = None) → D[source]

Updates column metadata with new metadata for column identified by column_index.

This stores column metadata under (ALL_ELEMENTS, column_index), at optionally at selector.

Parameters:
  • column_index (int) – Column index to update.
  • metadata (Dict) – A map of fields and values with metadata.
  • at (Selector) – Selector at which to assume tabular metadata.
  • source (primitive or Any) – DEPRECATED: argument ignored.
  • timestamp (datetime) – DEPRECATED: argument ignored.
Returns:

Updated column metadata.

Return type:

Metadata

class d3m.metadata.base.PrimitiveMetadata(metadata: Dict[str, Any] = None)[source]

Bases: d3m.metadata.base.Metadata

A class for metadata for primitives.

It checks all updates against primitive schema. Note that as such empty (just created) metadata object does not validate against the schema. If an instance is set on a primitive class, primitive’s metaclass logic will automatically link metadata object with the primitive class and generate required metadata.

clear(metadata: Dict[str, Any] = None, *, source: Any = None, timestamp: datetime.datetime = None) → P[source]

DEPRECATED: create a Metadata instance explicitly instead.

Creates and returns a new (clear) metadata object.

Parameters:
  • metadata (Dict[str, Any]) – Optional new initial metadata for the top-level of the value.
  • source (primitive or Any) – DEPRECATED: argument ignored.
  • timestamp (datetime) – DEPRECATED: argument ignored.
Returns:

New metadata object.

Return type:

Metadata

contribute_to_class(primitive: Any) → None[source]
get_hyperparams() → d3m.metadata.hyperparams.Hyperparams[source]
get_volumes() → Sequence[Dict][source]
query() → frozendict.FrozenOrderedDict[source]

Returns metadata for data pointed to with selector.

When querying using ALL_ELEMENTS means only metadata which has been set using ALL_ELEMENTS is returned.

Parameters:
  • selector (Tuple[Union[str, int, ALL_ELEMENTS]]) – A selector to query metadata for.
  • ignore_all_elements (bool) – By default, metadata from ALL_ELEMENTS is merged with metadata for an element itself. By setting this argument to True, this is disabled and just metadata from an element is returned.
  • remove_no_value (bool) – By default all NO_VALUE values are removed. If set to False, they are not removed.
Returns:

Metadata at a given selector.

Return type:

frozendict.FrozenOrderedDict

to_internal_json_structure() → Dict[source]

Converts metadata to a JSON-compatible structure.

The structure exposes how metadata is stored internally (metadata for ALL_ELEMENTS separate from metadata for individual elements) and can change in the future. This method exist for debugging purposes and to allow serialization of metadata. Use to_json_structure method if you want to access semantically valid representation of metadata.

Returns:A JSON-compatible list of dicts.
Return type:Sequence[Dict]
to_internal_simple_structure() → Dict[source]

Converts metadata to a simple structure, similar to JSON, but with values left as Python values.

The structure exposes how metadata is stored internally (metadata for ALL_ELEMENTS separate from metadata for individual elements) and can change in the future. This method exist for debugging purposes and to allow serialization of metadata. Use to_simple_structure method if you want to access semantically valid representation of metadata.

Returns:A list of dicts.
Return type:Sequence[Dict]
to_json_structure() → Dict[source]

Converts metadata to a JSON-compatible structure.

The output matches the output one obtain by using query method and is a semantically valid representation of metadata, but it does not matches how metadata is stored internally. To obtain that, you can use to_internal_json_structure method.

It does not make a JSON structure which can then be parsed back to reconstruct original metadata object. To obtain that, you can use to_internal_json_structure method.

Returns:A JSON-compatible list of dicts.
Return type:Sequence[Dict]
to_simple_structure() → Dict[source]

Converts metadata to a simple structure, similar to JSON, but with values left as Python values.

The output matches the output one obtain by using query method and is a semantically valid representation of metadata, but it does not matches how metadata is stored internally. To obtain that, you can use to_internal_simple_structure method.

It does not make a structure which can then be converted back to reconstruct original metadata object. To obtain that, you can use to_internal_simple_structure method.

Returns:A list of dicts.
Return type:Sequence[Dict]
update(metadata: Dict[str, Any], *, source: Any = None, timestamp: datetime.datetime = None) → P[source]

Updates metadata with new metadata for data pointed to with selector.

If value of any field is NO_VALUE, that field is deleted.

It returns a copy of this metadata object with new metadata applied.

Parameters:
  • selector (Tuple[Union[str, int, ALL_ELEMENTS]) – A selector pointing to data.
  • metadata (Dict) – A map of fields and values with metadata.
  • source (primitive or Any) – DEPRECATED: argument ignored.
  • timestamp (datetime) – DEPRECATED: argument ignored.
Returns:

Updated metadata.

Return type:

Metadata

class d3m.metadata.base.PrimitiveMethodKind[source]

Bases: d3m.utils.Enum

An enumeration.

OTHER = 'OTHER'[source]
PRODUCE = 'PRODUCE'[source]
class d3m.metadata.base.PrimitiveArgumentKind[source]

Bases: d3m.utils.Enum

An enumeration.

HYPERPARAMETER = 'HYPERPARAMETER'[source]
PIPELINE = 'PIPELINE'[source]
RUNTIME = 'RUNTIME'[source]
class d3m.metadata.base.PrimitiveInstallationType[source]

Bases: d3m.utils.Enum

An enumeration.

DOCKER = 'DOCKER'[source]
FILE = 'FILE'[source]
PIP = 'PIP'[source]
TGZ = 'TGZ'[source]
UBUNTU = 'UBUNTU'[source]
class d3m.metadata.base.PrimitiveAlgorithmType[source]

Bases: d3m.utils.Enum

An enumeration.

ACCURACY_SCORE = 'ACCURACY_SCORE'[source]
ADABOOST = 'ADABOOST'[source]
ADAPTIVE_ALGORITHM = 'ADAPTIVE_ALGORITHM'[source]
AGGREGATE_FUNCTION = 'AGGREGATE_FUNCTION'[source]
ALMEIDA_PINEDA_RECURRENT_BACKPROPAGATION = 'ALMEIDA_PINEDA_RECURRENT_BACKPROPAGATION'[source]
ALOPEX = 'ALOPEX'[source]
ALTERNATING_DECISION_TREE = 'ALTERNATING_DECISION_TREE'[source]
ANT_COLONY_OPTIMIZATION = 'ANT_COLONY_OPTIMIZATION'[source]
APPROXIMATE_DATA_AUGMENTATION = 'APPROXIMATE_DATA_AUGMENTATION'[source]
ARRAY_CONCATENATION = 'ARRAY_CONCATENATION'[source]
ARRAY_SLICING = 'ARRAY_SLICING'[source]
ASSOCIATION_RULE_LEARNING = 'ASSOCIATION_RULE_LEARNING'[source]
ASSOCIATIVE_NEURAL_NETWORK = 'ASSOCIATIVE_NEURAL_NETWORK'[source]
ATTRACTOR_NETWORK = 'ATTRACTOR_NETWORK'[source]
AUDIO_MIXING = 'AUDIO_MIXING'[source]
AUDIO_STREAM_MANIPULATION = 'AUDIO_STREAM_MANIPULATION'[source]
AUGMENTED_LAGRANGIAN_METHOD = 'AUGMENTED_LAGRANGIAN_METHOD'[source]
AUTOENCODER = 'AUTOENCODER'[source]
AUTOREGRESSIVE_INTEGRATED_MOVING_AVERAGE = 'AUTOREGRESSIVE_INTEGRATED_MOVING_AVERAGE'[source]
BACKWARD_DIFFERENCE_CODING = 'BACKWARD_DIFFERENCE_CODING'[source]
BAG_OF_WORDS_MODEL = 'BAG_OF_WORDS_MODEL'[source]
BATCH_NORMALIZATION = 'BATCH_NORMALIZATION'[source]
BAYESIAN_LINEAR_REGRESSION = 'BAYESIAN_LINEAR_REGRESSION'[source]
BAYESIAN_MODEL_AVERAGING = 'BAYESIAN_MODEL_AVERAGING'[source]
BAYESIAN_NETWORK = 'BAYESIAN_NETWORK'[source]
BAYESIAN_OPTIMIZATION = 'BAYESIAN_OPTIMIZATION'[source]
BELIEF_PROPAGATION = 'BELIEF_PROPAGATION'[source]
BERT = 'BERT'[source]
BINARY_CLASSIFICATION = 'BINARY_CLASSIFICATION'[source]
BIRCH = 'BIRCH'[source]
BOLTZMANN_MACHINE = 'BOLTZMANN_MACHINE'[source]
BOOSTING = 'BOOSTING'[source]
BOOTSTRAPPING = 'BOOTSTRAPPING'[source]
BOOTSTRAP_AGGREGATING = 'BOOTSTRAP_AGGREGATING'[source]
BRANCH_AND_BOUND = 'BRANCH_AND_BOUND'[source]
BRIER_SCORE = 'BRIER_SCORE'[source]
BROOKS_IYENGAR = 'BROOKS_IYENGAR'[source]
BROWNBOOST = 'BROWNBOOST'[source]
C45 = 'C45'[source]
C50 = 'C50'[source]
CANONICAL_CORRELATION_ANALYSIS = 'CANONICAL_CORRELATION_ANALYSIS'[source]
CASCADE_CORRELATION_NETWORK = 'CASCADE_CORRELATION_NETWORK'[source]
CASE_BASED_REASONING = 'CASE_BASED_REASONING'[source]
CATEGORY_ENCODER = 'CATEGORY_ENCODER'[source]
CLASSIFIER_CHAINS = 'CLASSIFIER_CHAINS'[source]
CN2 = 'CN2'[source]
COBWEB = 'COBWEB'[source]
COEFFICIENT_OF_DETERMINATION = 'COEFFICIENT_OF_DETERMINATION'[source]
COLOR_SPACE_CONVERSION = 'COLOR_SPACE_CONVERSION'[source]
COMMITTEE_MACHINE = 'COMMITTEE_MACHINE'[source]
COMPOSITIONAL_PATTERN_PRODUCING_NETWORK = 'COMPOSITIONAL_PATTERN_PRODUCING_NETWORK'[source]
COMPUTER_ALGEBRA = 'COMPUTER_ALGEBRA'[source]
CONDITIONAL_RANDOM_FIELD = 'CONDITIONAL_RANDOM_FIELD'[source]
CONTEXTUAL_BANDIT = 'CONTEXTUAL_BANDIT'[source]
CONVOLUTIONAL_NEURAL_NETWORK = 'CONVOLUTIONAL_NEURAL_NETWORK'[source]
CONVOLUTIONAL_NEURAL_NETWORK_LAYER = 'CONVOLUTIONAL_NEURAL_NETWORK_LAYER'[source]
COORDINATE_DESCENT = 'COORDINATE_DESCENT'[source]
CORRELATION_CLUSTERING = 'CORRELATION_CLUSTERING'[source]
CORTICAL_LEARNING = 'CORTICAL_LEARNING'[source]
COTRAINING = 'COTRAINING'[source]
CROSS_ENTROPY = 'CROSS_ENTROPY'[source]
CROSS_ENTROPY_METHOD = 'CROSS_ENTROPY_METHOD'[source]
CROSS_VALIDATION = 'CROSS_VALIDATION'[source]
CULTURAL_ALGORITHM = 'CULTURAL_ALGORITHM'[source]
DATA_CONVERSION = 'DATA_CONVERSION'[source]
DATA_DENORMALIZATION = 'DATA_DENORMALIZATION'[source]
DATA_MAPPING = 'DATA_MAPPING'[source]
DATA_NORMALIZATION = 'DATA_NORMALIZATION'[source]
DATA_PROFILING = 'DATA_PROFILING'[source]
DATA_RETRIEVAL = 'DATA_RETRIEVAL'[source]
DATA_SPLITTING = 'DATA_SPLITTING'[source]
DATA_STREAM_CLUSTERING = 'DATA_STREAM_CLUSTERING'[source]
DATA_STREAM_MINING = 'DATA_STREAM_MINING'[source]
DATA_STRUCTURE_ALIGNMENT = 'DATA_STRUCTURE_ALIGNMENT'[source]
DBSCAN = 'DBSCAN'[source]
DECISION_STUMP = 'DECISION_STUMP'[source]
DECISION_TREE = 'DECISION_TREE'[source]
DEEP_BELIEF_NETWORK = 'DEEP_BELIEF_NETWORK'[source]
DEEP_FEATURE_SYNTHESIS = 'DEEP_FEATURE_SYNTHESIS'[source]
DEEP_NEURAL_NETWORK = 'DEEP_NEURAL_NETWORK'[source]
DEINTERLACING = 'DEINTERLACING'[source]
DENSE_NEURAL_NETWORK_LAYER = 'DENSE_NEURAL_NETWORK_LAYER'[source]
DISCRETIZATION = 'DISCRETIZATION'[source]
DPLL = 'DPLL'[source]
DROPOUT = 'DROPOUT'[source]
DYNAMIC_NEURAL_NETWORK = 'DYNAMIC_NEURAL_NETWORK'[source]
DYNAMIC_TIME_WARPING = 'DYNAMIC_TIME_WARPING'[source]
EAGER_LEARNING = 'EAGER_LEARNING'[source]
ECHO_STATE_NETWORK = 'ECHO_STATE_NETWORK'[source]
ECLAT = 'ECLAT'[source]
EDGERANK = 'EDGERANK'[source]
ELASTIC_NET_REGULARIZATION = 'ELASTIC_NET_REGULARIZATION'[source]
ENCODE_BINARY = 'ENCODE_BINARY'[source]
ENCODE_ONE_HOT = 'ENCODE_ONE_HOT'[source]
ENCODE_ORDINAL = 'ENCODE_ORDINAL'[source]
ENCODE_UNARY = 'ENCODE_UNARY'[source]
ENSEMBLE_LEARNING = 'ENSEMBLE_LEARNING'[source]
EQUI_JOIN = 'EQUI_JOIN'[source]
EVOLUTIONARY_ACQUISITION_OF_NEURAL_TOPOLOGIES = 'EVOLUTIONARY_ACQUISITION_OF_NEURAL_TOPOLOGIES'[source]
EVOLUTIONARY_MULTIMODAL_OPTIMIZATION = 'EVOLUTIONARY_MULTIMODAL_OPTIMIZATION'[source]
EXPECTATION_MAXIMIZATION_ALGORITHM = 'EXPECTATION_MAXIMIZATION_ALGORITHM'[source]
EXTENSION_NEURAL_NETWORK = 'EXTENSION_NEURAL_NETWORK'[source]
EXTREME_LEARNING_MACHINE = 'EXTREME_LEARNING_MACHINE'[source]
F1_SCORE = 'F1_SCORE'[source]
FALSE_NEAREST_NEIGHBOR = 'FALSE_NEAREST_NEIGHBOR'[source]
FASTICA = 'FASTICA'[source]
FEATURE_SCALING = 'FEATURE_SCALING'[source]
FEEDFORWARD_NEURAL_NETWORK = 'FEEDFORWARD_NEURAL_NETWORK'[source]
FELLEGI_SUNTER_ALGORITHM = 'FELLEGI_SUNTER_ALGORITHM'[source]
FILE_MANIPULATION = 'FILE_MANIPULATION'[source]
FISHER_KERNEL = 'FISHER_KERNEL'[source]
FLATTEN_NEURAL_NETWORK_LAYER = 'FLATTEN_NEURAL_NETWORK_LAYER'[source]
FORWARD_ALGORITHM = 'FORWARD_ALGORITHM'[source]
FORWARD_BACKWARD_ALGORITHM = 'FORWARD_BACKWARD_ALGORITHM'[source]
FORWARD_DIFFERENCE_CODING = 'FORWARD_DIFFERENCE_CODING'[source]
FRANK_WOLFE_ALGORITHM = 'FRANK_WOLFE_ALGORITHM'[source]
FREQUENCY_TRANSFORM = 'FREQUENCY_TRANSFORM'[source]
FUZZY_CLUSTERING = 'FUZZY_CLUSTERING'[source]
GAUSSIAN_BLUR = 'GAUSSIAN_BLUR'[source]
GAUSSIAN_PROCESS = 'GAUSSIAN_PROCESS'[source]
GENERALIZED_HEBBIAN_ALGORITHM = 'GENERALIZED_HEBBIAN_ALGORITHM'[source]
GENERATIVE_TOPOGRAPHIC_MAP = 'GENERATIVE_TOPOGRAPHIC_MAP'[source]
GENETIC_ALGORITHM = 'GENETIC_ALGORITHM'[source]
GENETIC_ALGORITHM_FOR_RULE_SET_PRODUCTION = 'GENETIC_ALGORITHM_FOR_RULE_SET_PRODUCTION'[source]
GENETIC_PROGRAMMING = 'GENETIC_PROGRAMMING'[source]
GENETIC_SCALE_RECURRENT_NEURAL_NETWORK = 'GENETIC_SCALE_RECURRENT_NEURAL_NETWORK'[source]
GLOVE = 'GLOVE'[source]
GRADIENT_BOOSTING = 'GRADIENT_BOOSTING'[source]
GRADIENT_DESCENT = 'GRADIENT_DESCENT'[source]
GRAPHICAL_LASSO = 'GRAPHICAL_LASSO'[source]
GROWING_SELF_ORGANIZING_MAP = 'GROWING_SELF_ORGANIZING_MAP'[source]
HARD_CLUSTERING = 'HARD_CLUSTERING'[source]
HASHING = 'HASHING'[source]
HELMERT_CODING = 'HELMERT_CODING'[source]
HEURISTIC = 'HEURISTIC'[source]
HIDDEN_MARKOV_MODEL = 'HIDDEN_MARKOV_MODEL'[source]
HIDDEN_SEMI_MARKOV_MODEL = 'HIDDEN_SEMI_MARKOV_MODEL'[source]
HIERARCHICAL_CLUSTERING = 'HIERARCHICAL_CLUSTERING'[source]
HIERARCHICAL_TEMPORAL_MEMORY = 'HIERARCHICAL_TEMPORAL_MEMORY'[source]
HIGHER_ORDER_SINGULAR_VALUE_DECOMPOSITION = 'HIGHER_ORDER_SINGULAR_VALUE_DECOMPOSITION'[source]
HOLDOUT = 'HOLDOUT'[source]
HOLOGRAPHIC_ASSOCIATIVE_MEMORY = 'HOLOGRAPHIC_ASSOCIATIVE_MEMORY'[source]
HOPFIELD_NETWORK = 'HOPFIELD_NETWORK'[source]
HOSHEN_KOPELMAN_ALGORITHM = 'HOSHEN_KOPELMAN_ALGORITHM'[source]
HYPERNEAT = 'HYPERNEAT'[source]
HYPER_BASIS_FUNCTION_NETWORK = 'HYPER_BASIS_FUNCTION_NETWORK'[source]
ID3 = 'ID3'[source]
IDENTITY_FUNCTION = 'IDENTITY_FUNCTION'[source]
IMAGENET = 'IMAGENET'[source]
IMAGE_CROPPING = 'IMAGE_CROPPING'[source]
IMAGE_PADDING = 'IMAGE_PADDING'[source]
IMAGE_ROTATION = 'IMAGE_ROTATION'[source]
IMAGE_SCALING = 'IMAGE_SCALING'[source]
IMAGE_TRANSFORM = 'IMAGE_TRANSFORM'[source]
IMPUTATION = 'IMPUTATION'[source]
INDEPENDENT_COMPONENT_ANALYSIS = 'INDEPENDENT_COMPONENT_ANALYSIS'[source]
INFORMATION_ENTROPY = 'INFORMATION_ENTROPY'[source]
INFORMATION_FUZZY_NETWORKS = 'INFORMATION_FUZZY_NETWORKS'[source]
INFORMATION_THEORETIC_METAFEATURE_EXTRACTION = 'INFORMATION_THEORETIC_METAFEATURE_EXTRACTION'[source]
INSTANCE_BASED_LEARNING = 'INSTANCE_BASED_LEARNING'[source]
INSTANTANEOUSLY_TRAINED_NEURAL_NETWORKS = 'INSTANTANEOUSLY_TRAINED_NEURAL_NETWORKS'[source]
ISOMAP = 'ISOMAP'[source]
ITERATIVE_LABELING = 'ITERATIVE_LABELING'[source]
IVECTOR_EXTRACTION = 'IVECTOR_EXTRACTION'[source]
JACCARD_INDEX = 'JACCARD_INDEX'[source]
JUNCTION_TREE_ALGORITHM = 'JUNCTION_TREE_ALGORITHM'[source]
KERNEL_ADAPTIVE_FILTER = 'KERNEL_ADAPTIVE_FILTER'[source]
KERNEL_INDEPENDENT_COMPONENT_ANALYSIS = 'KERNEL_INDEPENDENT_COMPONENT_ANALYSIS'[source]
KERNEL_METHOD = 'KERNEL_METHOD'[source]
KERNEL_PERCEPTRON = 'KERNEL_PERCEPTRON'[source]
KERNEL_PRINCIPAL_COMPONENT_ANALYSIS = 'KERNEL_PRINCIPAL_COMPONENT_ANALYSIS'[source]
KERNEL_RANDOM_FOREST = 'KERNEL_RANDOM_FOREST'[source]
K_FOLD = 'K_FOLD'[source]
K_MEANS_CLUSTERING = 'K_MEANS_CLUSTERING'[source]
K_MEANS_PLUS_PLUS = 'K_MEANS_PLUS_PLUS'[source]
K_NEAREST_NEIGHBORS = 'K_NEAREST_NEIGHBORS'[source]
K_Q_FLATS = 'K_Q_FLATS'[source]
K_SVD = 'K_SVD'[source]
LANDMARKING_METAFEATURE_EXTRACTION = 'LANDMARKING_METAFEATURE_EXTRACTION'[source]
LARGE_MARGIN_NEAREST_NEIGHBOR = 'LARGE_MARGIN_NEAREST_NEIGHBOR'[source]
LASSO = 'LASSO'[source]
LATENT_DIRICHLET_ALLOCATION = 'LATENT_DIRICHLET_ALLOCATION'[source]
LATENT_SEMANTIC_ANALYSIS = 'LATENT_SEMANTIC_ANALYSIS'[source]
LEARNING_USING_PRIVILEGED_INFORMATION = 'LEARNING_USING_PRIVILEGED_INFORMATION'[source]
LEARNING_VECTOR_QUANTIZATION = 'LEARNING_VECTOR_QUANTIZATION'[source]
LEAST_SQUARES_SUPPORT_VECTOR_MACHINE = 'LEAST_SQUARES_SUPPORT_VECTOR_MACHINE'[source]
LEAVE_ONE_OUT = 'LEAVE_ONE_OUT'[source]
LIGHTGBM = 'LIGHTGBM'[source]
LIMITED_MEMORY_BFGS = 'LIMITED_MEMORY_BFGS'[source]
LINDE_BUZO_GRAY_ALGORITHM = 'LINDE_BUZO_GRAY_ALGORITHM'[source]
LINEAR_DISCRIMINANT_ANALYSIS = 'LINEAR_DISCRIMINANT_ANALYSIS'[source]
LINEAR_FILTER = 'LINEAR_FILTER'[source]
LINEAR_REGRESSION = 'LINEAR_REGRESSION'[source]
LOBPCG = 'LOBPCG'[source]
LOCAL_OUTLIER_FACTOR = 'LOCAL_OUTLIER_FACTOR'[source]
LOGISTIC_MODEL_TREE = 'LOGISTIC_MODEL_TREE'[source]
LOGISTIC_REGRESSION = 'LOGISTIC_REGRESSION'[source]
LOGITBOOST = 'LOGITBOOST'[source]
LONG_SHORT_TERM_MEMORY = 'LONG_SHORT_TERM_MEMORY'[source]
LOW_RANK_MATRIX_APPROXIMATIONS = 'LOW_RANK_MATRIX_APPROXIMATIONS'[source]
LPBOOST = 'LPBOOST'[source]
MAP = 'MAP'[source]
MARGIN_CLASSIFIER = 'MARGIN_CLASSIFIER'[source]
MARGIN_INFUSED_RELAXED_ALGORITHM = 'MARGIN_INFUSED_RELAXED_ALGORITHM'[source]
MARKOV_CHAIN = 'MARKOV_CHAIN'[source]
MARKOV_CHAIN_MONTE_CARLO = 'MARKOV_CHAIN_MONTE_CARLO'[source]
MARKOV_DECISION_PROCESS = 'MARKOV_DECISION_PROCESS'[source]
MARKOV_LOGIC_NETWORK = 'MARKOV_LOGIC_NETWORK'[source]
MARKOV_MODEL = 'MARKOV_MODEL'[source]
MARKOV_RANDOM_FIELD = 'MARKOV_RANDOM_FIELD'[source]
MAX_POOLING_NEURAL_NETWORK_LAYER = 'MAX_POOLING_NEURAL_NETWORK_LAYER'[source]
MEAN_ABSOLUTE_ERROR = 'MEAN_ABSOLUTE_ERROR'[source]
MEAN_SHIFT = 'MEAN_SHIFT'[source]
MEAN_SQUARED_ERROR = 'MEAN_SQUARED_ERROR'[source]
MEMETIC_ALGORITHM = 'MEMETIC_ALGORITHM'[source]
MEMORY_PREDICTION_FRAMEWORK = 'MEMORY_PREDICTION_FRAMEWORK'[source]
MERSENNE_TWISTER = 'MERSENNE_TWISTER'[source]
MFCC_FEATURE_EXTRACTION = 'MFCC_FEATURE_EXTRACTION'[source]
MINIMUM_REDUNDANCY_FEATURE_SELECTION = 'MINIMUM_REDUNDANCY_FEATURE_SELECTION'[source]
MINMAX_SCALER = 'MINMAX_SCALER'[source]
MIN_CONFLICTS_ALGORITHM = 'MIN_CONFLICTS_ALGORITHM'[source]
MM_ALGORITHM = 'MM_ALGORITHM'[source]
MODEL_BASED_METAFEATURE_EXTRACTION = 'MODEL_BASED_METAFEATURE_EXTRACTION'[source]
MODULAR_NEURAL_NETWORK = 'MODULAR_NEURAL_NETWORK'[source]
MORAVEC_CORNER_DETECTION_ALGORITHM = 'MORAVEC_CORNER_DETECTION_ALGORITHM'[source]
MOTION_COMPENSATION = 'MOTION_COMPENSATION'[source]
MULTICLASS_CLASSIFICATION = 'MULTICLASS_CLASSIFICATION'[source]
MULTILABEL_CLASSIFICATION = 'MULTILABEL_CLASSIFICATION'[source]
MULTILAYER_PERCEPTRON = 'MULTILAYER_PERCEPTRON'[source]
MULTINOMIAL_LOGISTIC_REGRESSION = 'MULTINOMIAL_LOGISTIC_REGRESSION'[source]
MULTINOMIAL_NAIVE_BAYES = 'MULTINOMIAL_NAIVE_BAYES'[source]
MULTIPLICATIVE_WEIGHT_UPDATE_METHOD = 'MULTIPLICATIVE_WEIGHT_UPDATE_METHOD'[source]
MULTIVARIATE_REGRESSION = 'MULTIVARIATE_REGRESSION'[source]
MULTI_ARMED_BANDIT = 'MULTI_ARMED_BANDIT'[source]
MUTUAL_INFORMATION = 'MUTUAL_INFORMATION'[source]
NAIVE_BAYES_CLASSIFIER = 'NAIVE_BAYES_CLASSIFIER'[source]
NEAREST_CENTROID_CLASSIFIER = 'NEAREST_CENTROID_CLASSIFIER'[source]
NEIGHBOURHOOD_COMPONENTS_ANALYSIS = 'NEIGHBOURHOOD_COMPONENTS_ANALYSIS'[source]
NEURAL_NETWORK_BACKPROPAGATION = 'NEURAL_NETWORK_BACKPROPAGATION'[source]
NEUROEVOLUTION_OF_AUGMENTED_TOPOLOGIES = 'NEUROEVOLUTION_OF_AUGMENTED_TOPOLOGIES'[source]
NEURO_FUZZY_NETWORK = 'NEURO_FUZZY_NETWORK'[source]
NOISE_REDUCTION = 'NOISE_REDUCTION'[source]
NONOVERLAPPING_COMMUNITY_DETECTION = 'NONOVERLAPPING_COMMUNITY_DETECTION'[source]
NORMAL_DISTRIBUTION = 'NORMAL_DISTRIBUTION'[source]
NUMERICAL_METHOD = 'NUMERICAL_METHOD'[source]
N_GRAM = 'N_GRAM'[source]
ONE_RULE = 'ONE_RULE'[source]
ONE_SHOT_ASSOCIATIVE_MEMORY = 'ONE_SHOT_ASSOCIATIVE_MEMORY'[source]
ONE_SHOT_LEARNING = 'ONE_SHOT_LEARNING'[source]
OPTICS_ALGORITHM = 'OPTICS_ALGORITHM'[source]
OPTIMISTIC_KNOWLEDGE_GRADIENT = 'OPTIMISTIC_KNOWLEDGE_GRADIENT'[source]
ORTHOGONAL_POLYNOMIAL_CODING = 'ORTHOGONAL_POLYNOMIAL_CODING'[source]
OVERLAPPING_CLUSTERING = 'OVERLAPPING_CLUSTERING'[source]
OVERLAPPING_COMMUNITY_DETECTION = 'OVERLAPPING_COMMUNITY_DETECTION'[source]
PACHINKO_ALLOCATION = 'PACHINKO_ALLOCATION'[source]
PAGERANK = 'PAGERANK'[source]
PARAMETRIC_TRAJECTORY_MODELING = 'PARAMETRIC_TRAJECTORY_MODELING'[source]
PARTIAL_LEAST_SQUARES_REGRESSION = 'PARTIAL_LEAST_SQUARES_REGRESSION'[source]
PARTICLE_SWARM_OPTIMIZATION = 'PARTICLE_SWARM_OPTIMIZATION'[source]
PASSIVE_AGGRESSIVE = 'PASSIVE_AGGRESSIVE'[source]
PERCEPTRON = 'PERCEPTRON'[source]
PHYSICAL_NEURAL_NETWORK = 'PHYSICAL_NEURAL_NETWORK'[source]
PIXELATION = 'PIXELATION'[source]
POLYNOMIAL_NEURAL_NETWORK = 'POLYNOMIAL_NEURAL_NETWORK'[source]
POLYNOMIAL_REGRESSION = 'POLYNOMIAL_REGRESSION'[source]
POPULATION_BASED_INCREMENTAL_LEARNING = 'POPULATION_BASED_INCREMENTAL_LEARNING'[source]
PREFRONTAL_CORTEX_BASAL_GANGLIA_WORKING_MEMORY = 'PREFRONTAL_CORTEX_BASAL_GANGLIA_WORKING_MEMORY'[source]
PRINCIPAL_COMPONENT_ANALYSIS = 'PRINCIPAL_COMPONENT_ANALYSIS'[source]
PROBABILISTIC_DATA_CLEANING = 'PROBABILISTIC_DATA_CLEANING'[source]
PROBABILISTIC_LATENT_SEMANTIC_ANALYSIS = 'PROBABILISTIC_LATENT_SEMANTIC_ANALYSIS'[source]
PROBABILISTIC_NEURAL_NETWORK = 'PROBABILISTIC_NEURAL_NETWORK'[source]
PRUNING = 'PRUNING'[source]
PSIPRED = 'PSIPRED'[source]
QUADRATIC_DISCRIMINANT_ANALYSIS = 'QUADRATIC_DISCRIMINANT_ANALYSIS'[source]
QUANTUM_NEURAL_NETWORK = 'QUANTUM_NEURAL_NETWORK'[source]
QUICKPROP = 'QUICKPROP'[source]
Q_LEARNING = 'Q_LEARNING'[source]
RADIAL_BASIS_FUNCTION_NETWORK = 'RADIAL_BASIS_FUNCTION_NETWORK'[source]
RANDOMIZED_WEIGHTED_MAJORITY_ALGORITHM = 'RANDOMIZED_WEIGHTED_MAJORITY_ALGORITHM'[source]
RANDOM_FOREST = 'RANDOM_FOREST'[source]
RANDOM_GRAPH = 'RANDOM_GRAPH'[source]
RANDOM_PROJECTION = 'RANDOM_PROJECTION'[source]
RANDOM_SUBSPACE_METHOD = 'RANDOM_SUBSPACE_METHOD'[source]
RANDOM_WALK = 'RANDOM_WALK'[source]
RANKBRAIN = 'RANKBRAIN'[source]
RANKING_SVM = 'RANKING_SVM'[source]
RAPIDLY_EXPLORING_RANDOM_TREE = 'RAPIDLY_EXPLORING_RANDOM_TREE'[source]
RECEIVER_OPERATING_CHARACTERISTIC = 'RECEIVER_OPERATING_CHARACTERISTIC'[source]
RECURRENT_NEURAL_NETWORK = 'RECURRENT_NEURAL_NETWORK'[source]
RECURSIVE_LEAST_SQUARES = 'RECURSIVE_LEAST_SQUARES'[source]
RECURSIVE_PARTITIONING = 'RECURSIVE_PARTITIONING'[source]
REGULARIZATION_BY_SPECTRAL_FILTERING = 'REGULARIZATION_BY_SPECTRAL_FILTERING'[source]
REGULARIZED_LEAST_SQUARES = 'REGULARIZED_LEAST_SQUARES'[source]
REGULATORY_FEEDBACK_NETWORK = 'REGULATORY_FEEDBACK_NETWORK'[source]
REINFORCE_ALGORITHM = 'REINFORCE_ALGORITHM'[source]
REJECTION_SAMPLING = 'REJECTION_SAMPLING'[source]
RELATIONAL_ALGEBRA = 'RELATIONAL_ALGEBRA'[source]
RELATIONAL_DATA_MINING = 'RELATIONAL_DATA_MINING'[source]
RELIEF = 'RELIEF'[source]
RESTRICTED_BOLTZMANN_MACHINE = 'RESTRICTED_BOLTZMANN_MACHINE'[source]
RETINANET = 'RETINANET'[source]
REVERSE_HELMERT_CODING = 'REVERSE_HELMERT_CODING'[source]
REVERSE_MONTE_CARLO = 'REVERSE_MONTE_CARLO'[source]
RIPPER = 'RIPPER'[source]
ROBUST_PRINCIPAL_COMPONENT_ANALYSIS = 'ROBUST_PRINCIPAL_COMPONENT_ANALYSIS'[source]
RPROP = 'RPROP'[source]
RULE_BASED_MACHINE_LEARNING = 'RULE_BASED_MACHINE_LEARNING'[source]
SAMPLE_MERGING = 'SAMPLE_MERGING'[source]
SAMPLE_SELECTION = 'SAMPLE_SELECTION'[source]
SELF_ORGANIZING_MAP = 'SELF_ORGANIZING_MAP'[source]
SEMIDEFINITE_EMBEDDING = 'SEMIDEFINITE_EMBEDDING'[source]
SIGNAL_DITHERING = 'SIGNAL_DITHERING'[source]
SIGNAL_ENERGY = 'SIGNAL_ENERGY'[source]
SIGNAL_TO_NOISE_RATIO = 'SIGNAL_TO_NOISE_RATIO'[source]
SIMULATED_ANNEALING = 'SIMULATED_ANNEALING'[source]
SINGULAR_VALUE_DECOMPOSITION = 'SINGULAR_VALUE_DECOMPOSITION'[source]
SMOOTHED_ANALYSIS = 'SMOOTHED_ANALYSIS'[source]
SOFTMAX_FUNCTION = 'SOFTMAX_FUNCTION'[source]
SOFT_CLUSTERING = 'SOFT_CLUSTERING'[source]
SPARSE_DICTIONARY_LEARNING = 'SPARSE_DICTIONARY_LEARNING'[source]
SPARSE_PCA = 'SPARSE_PCA'[source]
SPECTRAL_CLUSTERING = 'SPECTRAL_CLUSTERING'[source]
SPIKE_AND_SLAB_VARIABLE_SELECTION = 'SPIKE_AND_SLAB_VARIABLE_SELECTION'[source]
SPIKING_NEURAL_NETWORKS = 'SPIKING_NEURAL_NETWORKS'[source]
SPRUCE = 'SPRUCE'[source]
STATISTICAL_METAFEATURE_EXTRACTION = 'STATISTICAL_METAFEATURE_EXTRACTION'[source]
STATISTICAL_MOMENT_ANALYSIS = 'STATISTICAL_MOMENT_ANALYSIS'[source]
STOCHASTIC_CHAINS_WITH_MEMORY_OF_VARIABLE_LENGTH = 'STOCHASTIC_CHAINS_WITH_MEMORY_OF_VARIABLE_LENGTH'[source]
STOCHASTIC_GRADIENT_DESCENT = 'STOCHASTIC_GRADIENT_DESCENT'[source]
STOCHASTIC_NEURAL_NETWORK = 'STOCHASTIC_NEURAL_NETWORK'[source]
STRICT_PARTITIONING_CLUSTERING = 'STRICT_PARTITIONING_CLUSTERING'[source]
STRICT_PARTITIONING_CLUSTERING_WITH_OUTLIERS = 'STRICT_PARTITIONING_CLUSTERING_WITH_OUTLIERS'[source]
STRUCTURED_KNN = 'STRUCTURED_KNN'[source]
STRUCTURED_SPARSITY_REGULARIZATION = 'STRUCTURED_SPARSITY_REGULARIZATION'[source]
STRUCTURED_SUPPORT_VECTOR_MACHINE = 'STRUCTURED_SUPPORT_VECTOR_MACHINE'[source]
SUBSPACE_CLUSTERING = 'SUBSPACE_CLUSTERING'[source]
SUM_CODING = 'SUM_CODING'[source]
SUPER_RECURSIVE_ALGORITHM = 'SUPER_RECURSIVE_ALGORITHM'[source]
SUPPORT_VECTOR_MACHINE = 'SUPPORT_VECTOR_MACHINE'[source]
SYMBOLIC_REGRESSION = 'SYMBOLIC_REGRESSION'[source]
TFIDF = 'TFIDF'[source]
TIKHONOV_REGULARIZATION = 'TIKHONOV_REGULARIZATION'[source]
TIME_DELAY_NEURAL_NETWORK = 'TIME_DELAY_NEURAL_NETWORK'[source]
TRUNCATED_NEWTON_METHOD = 'TRUNCATED_NEWTON_METHOD'[source]
TRUNCATED_NORMAL_DISTRIBUTION = 'TRUNCATED_NORMAL_DISTRIBUTION'[source]
T_DISTRIBUTED_STOCHASTIC_NEIGHBOR_EMBEDDING = 'T_DISTRIBUTED_STOCHASTIC_NEIGHBOR_EMBEDDING'[source]
UNIFORM_DISTRIBUTION = 'UNIFORM_DISTRIBUTION'[source]
UNIFORM_TIME_SERIES_SEGMENTATION = 'UNIFORM_TIME_SERIES_SEGMENTATION'[source]
UNIT_WEIGHTED_REGRESSION = 'UNIT_WEIGHTED_REGRESSION'[source]
UNIVARIATE_REGRESSION = 'UNIVARIATE_REGRESSION'[source]
UNIVERSAL_PORTFOLIO_ALGORITHM = 'UNIVERSAL_PORTFOLIO_ALGORITHM'[source]
VARIABLE_ORDER_MARKOV_MODEL = 'VARIABLE_ORDER_MARKOV_MODEL'[source]
VARIATIONAL_BAYESIAN_METHODS = 'VARIATIONAL_BAYESIAN_METHODS'[source]
VARIATIONAL_MESSAGE_PASSING = 'VARIATIONAL_MESSAGE_PASSING'[source]
VECTORIZATION = 'VECTORIZATION'[source]
VECTOR_AUTOREGRESSION = 'VECTOR_AUTOREGRESSION'[source]
VECTOR_QUANTIZATION = 'VECTOR_QUANTIZATION'[source]
VERSION_SPACE_LEARNING = 'VERSION_SPACE_LEARNING'[source]
WAKE_SLEEP_ALGORITHM = 'WAKE_SLEEP_ALGORITHM'[source]
WEIGHTED_MAJORITY_ALGORITHM = 'WEIGHTED_MAJORITY_ALGORITHM'[source]
WINNOW = 'WINNOW'[source]
WORD2VEC = 'WORD2VEC'[source]
class d3m.metadata.base.PrimitiveFamily[source]

Bases: d3m.utils.Enum

An enumeration.

CLASSIFICATION = 'CLASSIFICATION'[source]
CLUSTERING = 'CLUSTERING'[source]
COLLABORATIVE_FILTERING = 'COLLABORATIVE_FILTERING'[source]
COMMUNITY_DETECTION = 'COMMUNITY_DETECTION'[source]
DATA_AUGMENTATION = 'DATA_AUGMENTATION'[source]
DATA_CLEANING = 'DATA_CLEANING'[source]
DATA_COMPRESSION = 'DATA_COMPRESSION'[source]
DATA_GENERATION = 'DATA_GENERATION'[source]
DATA_PREPROCESSING = 'DATA_PREPROCESSING'[source]
DATA_TRANSFORMATION = 'DATA_TRANSFORMATION'[source]
DATA_VALIDATION = 'DATA_VALIDATION'[source]
DATA_WRANGLING = 'DATA_WRANGLING'[source]
DIGITAL_IMAGE_PROCESSING = 'DIGITAL_IMAGE_PROCESSING'[source]
DIGITAL_SIGNAL_PROCESSING = 'DIGITAL_SIGNAL_PROCESSING'[source]
DIMENSIONALITY_REDUCTION = 'DIMENSIONALITY_REDUCTION'[source]
EVALUATION = 'EVALUATION'[source]
FEATURE_CONSTRUCTION = 'FEATURE_CONSTRUCTION'[source]
FEATURE_EXTRACTION = 'FEATURE_EXTRACTION'[source]
FEATURE_SELECTION = 'FEATURE_SELECTION'[source]
GRAPH_CLUSTERING = 'GRAPH_CLUSTERING'[source]
GRAPH_MATCHING = 'GRAPH_MATCHING'[source]
LAYER = 'LAYER'[source]
LEARNER = 'LEARNER'[source]
LOSS_FUNCTION = 'LOSS_FUNCTION'[source]
METALEARNING = 'METALEARNING'[source]
NATURAL_LANGUAGE_PROCESSING = 'NATURAL_LANGUAGE_PROCESSING'[source]
NORMALIZATION = 'NORMALIZATION'[source]
OBJECT_DETECTION = 'OBJECT_DETECTION'[source]
OPERATOR = 'OPERATOR'[source]
REGRESSION = 'REGRESSION'[source]
REMOTE_SENSING = 'REMOTE_SENSING'[source]
SCHEMA_DISCOVERY = 'SCHEMA_DISCOVERY'[source]
SEMISUPERVISED_CLASSIFICATION = 'SEMISUPERVISED_CLASSIFICATION'[source]
SEMISUPERVISED_REGRESSION = 'SEMISUPERVISED_REGRESSION'[source]
SIMILARITY_MODELING = 'SIMILARITY_MODELING'[source]
TIME_SERIES_CLASSIFICATION = 'TIME_SERIES_CLASSIFICATION'[source]
TIME_SERIES_EMBEDDING = 'TIME_SERIES_EMBEDDING'[source]
TIME_SERIES_FORECASTING = 'TIME_SERIES_FORECASTING'[source]
TIME_SERIES_SEGMENTATION = 'TIME_SERIES_SEGMENTATION'[source]
VERTEX_CLASSIFICATION = 'VERTEX_CLASSIFICATION'[source]
VERTEX_NOMINATION = 'VERTEX_NOMINATION'[source]
VIDEO_PROCESSING = 'VIDEO_PROCESSING'[source]
class d3m.metadata.base.PrimitivePrecondition[source]

Bases: d3m.utils.Enum

An enumeration.

NO_CATEGORICAL_VALUES = 'NO_CATEGORICAL_VALUES'[source]
NO_CONTINUOUS_VALUES = 'NO_CONTINUOUS_VALUES'[source]
NO_JAGGED_VALUES = 'NO_JAGGED_VALUES'[source]
NO_MISSING_VALUES = 'NO_MISSING_VALUES'[source]
NO_NEGATIVE_VALUES = 'NO_NEGATIVE_VALUES'[source]
NO_NESTED_VALUES = 'NO_NESTED_VALUES'[source]
class d3m.metadata.base.PrimitiveEffect[source]

Bases: d3m.utils.Enum

An enumeration.

NO_CATEGORICAL_VALUES = 'NO_CATEGORICAL_VALUES'[source]
NO_CONTINUOUS_VALUES = 'NO_CONTINUOUS_VALUES'[source]
NO_JAGGED_VALUES = 'NO_JAGGED_VALUES'[source]
NO_MISSING_VALUES = 'NO_MISSING_VALUES'[source]
NO_NEGATIVE_VALUES = 'NO_NEGATIVE_VALUES'[source]
NO_NESTED_VALUES = 'NO_NESTED_VALUES'[source]
class d3m.metadata.base.ForeignKeyType[source]

Bases: d3m.utils.Enum

An enumeration.

COLUMN = 'COLUMN'[source]
EDGE_ATTRIBUTE = 'EDGE_ATTRIBUTE'[source]
NODE_ATTRIBUTE = 'NODE_ATTRIBUTE'[source]
RESOURCE = 'RESOURCE'[source]
class d3m.metadata.base.Context[source]

Bases: d3m.utils.Enum

An enumeration.

EVALUATION = 'EVALUATION'[source]
PRETRAINING = 'PRETRAINING'[source]
PRODUCTION = 'PRODUCTION'[source]
TESTING = 'TESTING'[source]
class d3m.metadata.base.PipelineRunPhase[source]

Bases: d3m.utils.Enum

An enumeration.

FIT = 'FIT'[source]
PRODUCE = 'PRODUCE'[source]
class d3m.metadata.base.PipelineStepType[source]

Bases: d3m.utils.Enum

An enumeration.

PLACEHOLDER = 'PLACEHOLDER'[source]
PRIMITIVE = 'PRIMITIVE'[source]
SUBPIPELINE = 'SUBPIPELINE'[source]
d3m.metadata.base.PipelineRunStatusState[source]

alias of d3m.metadata.base.StatusState

class d3m.metadata.base.ArgumentType[source]

Bases: d3m.utils.Enum

An enumeration.

CONTAINER = 'CONTAINER'[source]
DATA = 'DATA'[source]
PRIMITIVE = 'PRIMITIVE'[source]
VALUE = 'VALUE'[source]