Class: EnsmallenRunConfig
All params used by ensmallen's csv_reader
URI: https://w3id.org/neat-ml/:EnsmallenRunConfig
Slots
| Name | Range | Cardinality | Description | Info |
|---|---|---|---|---|
| node_type_path | string | 0..1 | The path to the file with the unique node type names. | . |
| node_type_list_separator | string | 0..1 | The separator to use for the node types file. | . |
| node_types_column_number | integer | 0..1 | Node type column number. | . |
| node_types_column | string | 0..1 | Node type column name. | . |
| node_types_ids_column_number | integer | 0..1 | Node type ID column number. | . |
| node_types_ids_column | string | 0..1 | Node type ID column name. | . |
| node_types_number | integer | 0..1 | The number of the unique node types. This will be used in order to allocate the correct size for the data structure. | . |
| numeric_node_type_ids | boolean | 0..1 | Whether the node type names should be loaded as numeric values, i.e. casted from string to a numeric representation. | . |
| minimum_node_type_id | integer | 0..1 | The minimum node type ID to be used when using numeric node type IDs. | . |
| node_type_list_header | boolean | 0..1 | Whether the node type file has an header. | . |
| node_type_list_support_balanced_quotes | boolean | 0..1 | Whether to support balanced quotes. | . |
| node_type_list_rows_to_skip | integer | 0..1 | The number of lines to skip in the node types file. The header is already skipped if it has been specified that the file has an header. | . |
| node_type_list_is_correct | boolean | 0..1 | Whether the node types file can be assumed to be correct, i.e. does not have something wrong in it. If this parameter is passed as true on a malformed file, the constructor will crash. | . |
| node_type_list_max_rows_number | integer | 0..1 | The maximum number of lines to be loaded from the node types file. | . |
| node_type_list_comment_symbol | string | 0..1 | The comment symbol to skip lines in the node types file. Lines starting with this symbol will be skipped. | . |
| load_node_type_list_in_parallel | boolean | 0..1 | Whether to load the node type list in parallel. Note that when loading in parallel, the internal order of the node type IDs may result changed across different iterations. We are working to get this to be stable. | . |
| node_path | string | 0..1 | The path to the file with the unique node names. | . |
| node_list_separator | string | 0..1 | The separator to use for the nodes file. Note that if this is not provided, one will be automatically detected among the following - comma, semi-column, tab and space. | . |
| node_list_header | boolean | 0..1 | Whether the nodes file has an header. | . |
| node_list_support_balanced_quotes | boolean | 0..1 | Whether to support balanced quotes. | . |
| node_list_rows_to_skip | integer | 0..1 | Number of rows to skip in the node list file. | . |
| node_list_is_correct | boolean | 0..1 | Whether the nodes file can be assumed to be correct, i.e. does not have something wrong in it. If this parameter is passed as true on a malformed file, the constructor will crash. | . |
| node_list_max_rows_number | integer | 0..1 | The maximum number of lines to be loaded from the nodes file. | . |
| node_list_comment_symbol | string | 0..1 | The comment symbol to skip lines in the nodes file. Lines starting with this symbol will be skipped. | . |
| default_node_type | string | 0..1 | The node type to be used when the node type for a given node in the node file is None. | . |
| nodes_column_number | integer | 0..1 | The number of the column of the node file from where to load the node names. | . |
| nodes_column | string | 0..1 | The name of the column of the node file from where to load the node names. | . |
| node_types_separator | string | 0..1 | The node types separator. | . |
| node_list_node_types_column_number | integer | 0..1 | The number of the column of the node file from where to load the node types. | . |
| node_list_node_types_column | string | 0..1 | The name of the column of the node file from where to load the node types. | . |
| node_ids_column | string | 0..1 | The name of the column of the node file from where to load the node IDs. | . |
| node_ids_column_number | integer | 0..1 | The number of the column of the node file from where to load the node IDs | . |
| nodes_number | integer | 0..1 | The expected number of nodes. Note that this must be the EXACT number of nodes in the graph. | . |
| minimum_node_id | integer | 0..1 | The minimum node ID to be used, when loading the node IDs as numerical. | . |
| numeric_node_ids | boolean | 0..1 | Whether to load the numeric node IDs as numeric. | . |
| node_list_numeric_node_type_ids | boolean | 0..1 | Whether to load the node types IDs in the node file to be numeric. | . |
| skip_node_types_if_unavailable | boolean | 0..1 | Whether to skip the node types without raising an error if these are unavailable. | . |
| load_node_list_in_parallel | boolean | 0..1 | Whether to load the node list in parallel. When loading in parallel, without node IDs, the nodes may not be loaded in a deterministic order. | . |
| edge_type_path | string | 0..1 | The path to the file with the unique edge type names. | . |
| edge_types_column_number | integer | 0..1 | The number of the column of the edge types file from where to load the edge types. | . |
| edge_types_column | string | 0..1 | The name of the column of the edge types file from where to load the edge types. | . |
| edge_types_number | integer | 0..1 | The number of the unique edge types. This will be used in order to allocate the correct size for the data structure. | . |
| numeric_edge_type_ids | boolean | 0..1 | Whether the edge type names should be loaded as numeric values, i.e. casted from string to a numeric representation. | . |
| minimum_edge_type_id | integer | 0..1 | The minimum edge type ID to be used when using numeric edge type IDs. | . |
| edge_type_list_separator | string | 0..1 | The separator to use for the edge type list. Note that, if None is provided, one will be attempted to be detected automatically between ';', ',', tab or space. | . |
| edge_type_list_header | boolean | 0..1 | Whether the edge type file has an header. | . |
| edge_type_list_support_balanced_quotes | boolean | 0..1 | Whether to support balanced quotes while reading the edge type list. | . |
| edge_type_list_rows_to_skip | integer | 0..1 | Number of rows to skip in the edge type list file. | . |
| edge_type_list_is_correct | boolean | 0..1 | Whether the edge types file can be assumed to be correct, i.e. does not have something wrong in it. If this parameter is passed as true on a malformed file, the constructor will crash. | . |
| edge_type_list_max_rows_number | integer | 0..1 | The maximum number of lines to be loaded from the edge types file. | . |
| edge_type_list_comment_symbol | string | 0..1 | The comment symbol to skip lines in the edge types file. Lines starting with this symbol will be skipped. | . |
| load_edge_type_list_in_parallel | boolean | 0..1 | Whether to load the edge type list in parallel. When loading in parallel, without edge type IDs, the edge types may not be loaded in a deterministic order. | . |
| edge_path | string | 0..1 | The path to the file with the edge list. | . |
| edge_list_separator | string | 0..1 | The separator to use for the edge list. Note that, if None is provided, one will be attempted to be detected automatically between ';', ',', tab or space. | . |
| edge_list_header | boolean | 0..1 | Whether the edges file has an header. | . |
| edge_list_support_balanced_quotes | boolean | 0..1 | Whether to support balanced quotes while reading the edge list. | . |
| edge_list_rows_to_skip | integer | 0..1 | Number of rows to skip in the edge list file. | . |
| sources_column_number | integer | 0..1 | The number of the column of the edges file from where to load the source nodes. | . |
| sources_column | string | 0..1 | The name of the column of the edges file from where to load the source nodes. | . |
| destinations_column_number | integer | 0..1 | The number of the column of the edges file from where to load the destinaton nodes. | . |
| destinations_column | string | 0..1 | The name of the column of the edges file from where to load the destinaton nodes. | . |
| edge_list_edge_types_column_number | integer | 0..1 | The number of the column of the edges file from where to load the edge types. | . |
| edge_list_edge_types_column | string | 0..1 | The name of the column of the edges file from where to load the edge types. | . |
| default_edge_type | string | 0..1 | The edge type to be used when the edge type for a given edge in the edge file is None. | . |
| weights_column_number | integer | 0..1 | The number of the column of the edges file from where to load the edge weights. | . |
| weights_column | string | 0..1 | The name of the column of the edges file from where to load the edge weights. | . |
| default_weight | float | 0..1 | The edge weight to be used when the edge weight for a given edge in the edge file is None. | . |
| edge_ids_column | string | 0..1 | The name of the column of the edges file from where to load the edge IDs. | . |
| edge_ids_column_number | integer | 0..1 | The number of the column of the edges file from where to load the edge IDs. | . |
| edge_list_numeric_edge_type_ids | boolean | 0..1 | Whether to load the edge type IDs as numeric from the edge list. | . |
| edge_list_numeric_node_ids | boolean | 0..1 | Whether to load the edge node IDs as numeric from the edge list. | . |
| skip_weights_if_unavailable | boolean | 0..1 | Whether to skip the weights without raising an error if these are unavailable. | . |
| skip_edge_types_if_unavailable | boolean | 0..1 | Whether to skip the edge types without raising an error if these are unavailable. | . |
| edge_list_is_complete | boolean | 0..1 | Whether to consider the edge list as complete, i.e. the edges are presented in both directions when loading an undirected graph. | . |
| edge_list_may_contain_duplicates | boolean | 0..1 | Whether the edge list may contain duplicates. If the edge list surely DOES NOT contain duplicates, a validation step may be skipped. By default, it is assumed that the edge list may contain duplicates. | . |
| edge_list_is_sorted | boolean | 0..1 | Whether the edge list is sorted. Note that a sorted edge list has the minimal memory peak, but requires the nodes number and the edges number. | . |
| edge_list_is_correct | boolean | 0..1 | Whether the edges file can be assumed to be correct, i.e. does not have something wrong in it. If this parameter is passed as true on a malformed file, the constructor will crash. | . |
| edge_list_max_rows_number | integer | 0..1 | The maximum number of lines to be loaded from the edges file. | . |
| edge_list_comment_symbol | string | 0..1 | The comment symbol to skip lines in the edges file. Lines starting with this symbol will be skipped. | . |
| edges_number | integer | 0..1 | The expected number of edges. Note that this must be the EXACT number of edges in the graph. | . |
| load_edge_list_in_parallel | boolean | 0..1 | Whether to load the edge list in parallel. Note that, if the edge IDs indices are not given, it is NOT possible to load a sorted edge list. Similarly, when loading in parallel, without edge IDs, the edges may not be loaded in a deterministic order. | . |
| verbose | boolean | 0..1 | Whether to show a loading bar while reading the files. Note that, if parallel loading is enabled, loading bars will not be showed because they are a synchronization bottleneck. | . |
| may_have_singletons | boolean | 0..1 | Whether the graph may be expected to have singleton nodes. If it is said that it surely DOES NOT have any, it will allow for some speedups and lower mempry peaks. | . |
| may_have_singleton_with_selfloops | boolean | 0..1 | Whether the graph may be expected to have singleton nodes with selfloops. If it is said that it surely DOES NOT have any, it will allow for some speedups and lower mempry peaks. | . |
| directed | boolean | 0..1 | Whether to load the graph as directed or undirected. | . |
| name | string | 0..1 | None | . |
Usages
| used by | used in | type | used |
|---|---|---|---|
| GraphDataConfiguration | graph | range | EnsmallenRunConfig |
Identifier and Mapping Information
LinkML Specification
Direct
name: EnsmallenRunConfig
description: All params used by ensmallen's csv_reader
from_schema: https://w3id.org/neat-ml
attributes:
node_type_path:
name: node_type_path
description: The path to the file with the unique node type names.
from_schema: https://w3id.org/neat-ml
node_type_list_separator:
name: node_type_list_separator
description: The separator to use for the node types file.
from_schema: https://w3id.org/neat-ml
node_types_column_number:
name: node_types_column_number
description: Node type column number.
from_schema: https://w3id.org/neat-ml
range: integer
node_types_column:
name: node_types_column
description: Node type column name.
from_schema: https://w3id.org/neat-ml
node_types_ids_column_number:
name: node_types_ids_column_number
description: Node type ID column number.
from_schema: https://w3id.org/neat-ml
range: integer
node_types_ids_column:
name: node_types_ids_column
description: Node type ID column name.
from_schema: https://w3id.org/neat-ml
node_types_number:
name: node_types_number
description: The number of the unique node types. This will be used in order to
allocate the correct size for the data structure.
from_schema: https://w3id.org/neat-ml
range: integer
numeric_node_type_ids:
name: numeric_node_type_ids
description: Whether the node type names should be loaded as numeric values, i.e.
casted from string to a numeric representation.
from_schema: https://w3id.org/neat-ml
range: boolean
minimum_node_type_id:
name: minimum_node_type_id
description: The minimum node type ID to be used when using numeric node type
IDs.
from_schema: https://w3id.org/neat-ml
range: integer
node_type_list_header:
name: node_type_list_header
description: Whether the node type file has an header.
from_schema: https://w3id.org/neat-ml
range: boolean
node_type_list_support_balanced_quotes:
name: node_type_list_support_balanced_quotes
description: Whether to support balanced quotes.
from_schema: https://w3id.org/neat-ml
range: boolean
node_type_list_rows_to_skip:
name: node_type_list_rows_to_skip
description: The number of lines to skip in the node types file. The header is
already skipped if it has been specified that the file has an header.
from_schema: https://w3id.org/neat-ml
range: integer
node_type_list_is_correct:
name: node_type_list_is_correct
description: Whether the node types file can be assumed to be correct, i.e. does
not have something wrong in it. If this parameter is passed as true on a malformed
file, the constructor will crash.
from_schema: https://w3id.org/neat-ml
range: boolean
node_type_list_max_rows_number:
name: node_type_list_max_rows_number
description: The maximum number of lines to be loaded from the node types file.
from_schema: https://w3id.org/neat-ml
range: integer
node_type_list_comment_symbol:
name: node_type_list_comment_symbol
description: The comment symbol to skip lines in the node types file. Lines starting
with this symbol will be skipped.
from_schema: https://w3id.org/neat-ml
load_node_type_list_in_parallel:
name: load_node_type_list_in_parallel
description: Whether to load the node type list in parallel. Note that when loading
in parallel, the internal order of the node type IDs may result changed across
different iterations. We are working to get this to be stable.
from_schema: https://w3id.org/neat-ml
range: boolean
node_path:
name: node_path
description: The path to the file with the unique node names.
from_schema: https://w3id.org/neat-ml
node_list_separator:
name: node_list_separator
description: The separator to use for the nodes file. Note that if this is not
provided, one will be automatically detected among the following - comma, semi-column,
tab and space.
from_schema: https://w3id.org/neat-ml
node_list_header:
name: node_list_header
description: Whether the nodes file has an header.
from_schema: https://w3id.org/neat-ml
range: boolean
node_list_support_balanced_quotes:
name: node_list_support_balanced_quotes
description: Whether to support balanced quotes.
from_schema: https://w3id.org/neat-ml
range: boolean
node_list_rows_to_skip:
name: node_list_rows_to_skip
description: Number of rows to skip in the node list file.
from_schema: https://w3id.org/neat-ml
range: integer
node_list_is_correct:
name: node_list_is_correct
description: Whether the nodes file can be assumed to be correct, i.e. does not
have something wrong in it. If this parameter is passed as true on a malformed
file, the constructor will crash.
from_schema: https://w3id.org/neat-ml
range: boolean
node_list_max_rows_number:
name: node_list_max_rows_number
description: The maximum number of lines to be loaded from the nodes file.
from_schema: https://w3id.org/neat-ml
range: integer
node_list_comment_symbol:
name: node_list_comment_symbol
description: The comment symbol to skip lines in the nodes file. Lines starting
with this symbol will be skipped.
from_schema: https://w3id.org/neat-ml
default_node_type:
name: default_node_type
description: The node type to be used when the node type for a given node in the
node file is None.
from_schema: https://w3id.org/neat-ml
nodes_column_number:
name: nodes_column_number
description: The number of the column of the node file from where to load the
node names.
from_schema: https://w3id.org/neat-ml
range: integer
nodes_column:
name: nodes_column
description: The name of the column of the node file from where to load the node
names.
from_schema: https://w3id.org/neat-ml
node_types_separator:
name: node_types_separator
description: The node types separator.
from_schema: https://w3id.org/neat-ml
node_list_node_types_column_number:
name: node_list_node_types_column_number
description: The number of the column of the node file from where to load the
node types.
from_schema: https://w3id.org/neat-ml
range: integer
node_list_node_types_column:
name: node_list_node_types_column
description: The name of the column of the node file from where to load the node
types.
from_schema: https://w3id.org/neat-ml
node_ids_column:
name: node_ids_column
description: The name of the column of the node file from where to load the node
IDs.
from_schema: https://w3id.org/neat-ml
node_ids_column_number:
name: node_ids_column_number
description: The number of the column of the node file from where to load the
node IDs
from_schema: https://w3id.org/neat-ml
range: integer
nodes_number:
name: nodes_number
description: The expected number of nodes. Note that this must be the EXACT number
of nodes in the graph.
from_schema: https://w3id.org/neat-ml
range: integer
minimum_node_id:
name: minimum_node_id
description: The minimum node ID to be used, when loading the node IDs as numerical.
from_schema: https://w3id.org/neat-ml
range: integer
numeric_node_ids:
name: numeric_node_ids
description: Whether to load the numeric node IDs as numeric.
from_schema: https://w3id.org/neat-ml
range: boolean
node_list_numeric_node_type_ids:
name: node_list_numeric_node_type_ids
description: Whether to load the node types IDs in the node file to be numeric.
from_schema: https://w3id.org/neat-ml
range: boolean
skip_node_types_if_unavailable:
name: skip_node_types_if_unavailable
description: Whether to skip the node types without raising an error if these
are unavailable.
from_schema: https://w3id.org/neat-ml
range: boolean
load_node_list_in_parallel:
name: load_node_list_in_parallel
description: Whether to load the node list in parallel. When loading in parallel,
without node IDs, the nodes may not be loaded in a deterministic order.
from_schema: https://w3id.org/neat-ml
range: boolean
edge_type_path:
name: edge_type_path
description: The path to the file with the unique edge type names.
from_schema: https://w3id.org/neat-ml
edge_types_column_number:
name: edge_types_column_number
description: The number of the column of the edge types file from where to load
the edge types.
from_schema: https://w3id.org/neat-ml
range: integer
edge_types_column:
name: edge_types_column
description: The name of the column of the edge types file from where to load
the edge types.
from_schema: https://w3id.org/neat-ml
edge_types_number:
name: edge_types_number
description: The number of the unique edge types. This will be used in order to
allocate the correct size for the data structure.
from_schema: https://w3id.org/neat-ml
range: integer
numeric_edge_type_ids:
name: numeric_edge_type_ids
description: Whether the edge type names should be loaded as numeric values, i.e.
casted from string to a numeric representation.
from_schema: https://w3id.org/neat-ml
range: boolean
minimum_edge_type_id:
name: minimum_edge_type_id
description: The minimum edge type ID to be used when using numeric edge type
IDs.
from_schema: https://w3id.org/neat-ml
range: integer
edge_type_list_separator:
name: edge_type_list_separator
description: The separator to use for the edge type list. Note that, if None is
provided, one will be attempted to be detected automatically between ';', ',',
tab or space.
from_schema: https://w3id.org/neat-ml
edge_type_list_header:
name: edge_type_list_header
description: Whether the edge type file has an header.
from_schema: https://w3id.org/neat-ml
range: boolean
edge_type_list_support_balanced_quotes:
name: edge_type_list_support_balanced_quotes
description: Whether to support balanced quotes while reading the edge type list.
from_schema: https://w3id.org/neat-ml
range: boolean
edge_type_list_rows_to_skip:
name: edge_type_list_rows_to_skip
description: Number of rows to skip in the edge type list file.
from_schema: https://w3id.org/neat-ml
range: integer
edge_type_list_is_correct:
name: edge_type_list_is_correct
description: Whether the edge types file can be assumed to be correct, i.e. does
not have something wrong in it. If this parameter is passed as true on a malformed
file, the constructor will crash.
from_schema: https://w3id.org/neat-ml
range: boolean
edge_type_list_max_rows_number:
name: edge_type_list_max_rows_number
description: The maximum number of lines to be loaded from the edge types file.
from_schema: https://w3id.org/neat-ml
range: integer
edge_type_list_comment_symbol:
name: edge_type_list_comment_symbol
description: The comment symbol to skip lines in the edge types file. Lines starting
with this symbol will be skipped.
from_schema: https://w3id.org/neat-ml
load_edge_type_list_in_parallel:
name: load_edge_type_list_in_parallel
description: Whether to load the edge type list in parallel. When loading in parallel,
without edge type IDs, the edge types may not be loaded in a deterministic order.
from_schema: https://w3id.org/neat-ml
range: boolean
edge_path:
name: edge_path
description: The path to the file with the edge list.
from_schema: https://w3id.org/neat-ml
edge_list_separator:
name: edge_list_separator
description: The separator to use for the edge list. Note that, if None is provided,
one will be attempted to be detected automatically between ';', ',', tab or
space.
from_schema: https://w3id.org/neat-ml
edge_list_header:
name: edge_list_header
description: Whether the edges file has an header.
from_schema: https://w3id.org/neat-ml
range: boolean
edge_list_support_balanced_quotes:
name: edge_list_support_balanced_quotes
description: Whether to support balanced quotes while reading the edge list.
from_schema: https://w3id.org/neat-ml
range: boolean
edge_list_rows_to_skip:
name: edge_list_rows_to_skip
description: Number of rows to skip in the edge list file.
from_schema: https://w3id.org/neat-ml
range: integer
sources_column_number:
name: sources_column_number
description: The number of the column of the edges file from where to load the
source nodes.
from_schema: https://w3id.org/neat-ml
range: integer
sources_column:
name: sources_column
description: The name of the column of the edges file from where to load the source
nodes.
from_schema: https://w3id.org/neat-ml
destinations_column_number:
name: destinations_column_number
description: The number of the column of the edges file from where to load the
destinaton nodes.
from_schema: https://w3id.org/neat-ml
range: integer
destinations_column:
name: destinations_column
description: The name of the column of the edges file from where to load the destinaton
nodes.
from_schema: https://w3id.org/neat-ml
edge_list_edge_types_column_number:
name: edge_list_edge_types_column_number
description: The number of the column of the edges file from where to load the
edge types.
from_schema: https://w3id.org/neat-ml
range: integer
edge_list_edge_types_column:
name: edge_list_edge_types_column
description: The name of the column of the edges file from where to load the edge
types.
from_schema: https://w3id.org/neat-ml
default_edge_type:
name: default_edge_type
description: The edge type to be used when the edge type for a given edge in the
edge file is None.
from_schema: https://w3id.org/neat-ml
weights_column_number:
name: weights_column_number
description: The number of the column of the edges file from where to load the
edge weights.
from_schema: https://w3id.org/neat-ml
range: integer
weights_column:
name: weights_column
description: The name of the column of the edges file from where to load the edge
weights.
from_schema: https://w3id.org/neat-ml
default_weight:
name: default_weight
description: The edge weight to be used when the edge weight for a given edge
in the edge file is None.
from_schema: https://w3id.org/neat-ml
range: float
edge_ids_column:
name: edge_ids_column
description: The name of the column of the edges file from where to load the edge
IDs.
from_schema: https://w3id.org/neat-ml
edge_ids_column_number:
name: edge_ids_column_number
description: The number of the column of the edges file from where to load the
edge IDs.
from_schema: https://w3id.org/neat-ml
range: integer
edge_list_numeric_edge_type_ids:
name: edge_list_numeric_edge_type_ids
description: Whether to load the edge type IDs as numeric from the edge list.
from_schema: https://w3id.org/neat-ml
range: boolean
edge_list_numeric_node_ids:
name: edge_list_numeric_node_ids
description: Whether to load the edge node IDs as numeric from the edge list.
from_schema: https://w3id.org/neat-ml
range: boolean
skip_weights_if_unavailable:
name: skip_weights_if_unavailable
description: Whether to skip the weights without raising an error if these are
unavailable.
from_schema: https://w3id.org/neat-ml
range: boolean
skip_edge_types_if_unavailable:
name: skip_edge_types_if_unavailable
description: Whether to skip the edge types without raising an error if these
are unavailable.
from_schema: https://w3id.org/neat-ml
range: boolean
edge_list_is_complete:
name: edge_list_is_complete
description: Whether to consider the edge list as complete, i.e. the edges are
presented in both directions when loading an undirected graph.
from_schema: https://w3id.org/neat-ml
range: boolean
edge_list_may_contain_duplicates:
name: edge_list_may_contain_duplicates
description: Whether the edge list may contain duplicates. If the edge list surely
DOES NOT contain duplicates, a validation step may be skipped. By default, it
is assumed that the edge list may contain duplicates.
from_schema: https://w3id.org/neat-ml
range: boolean
edge_list_is_sorted:
name: edge_list_is_sorted
description: Whether the edge list is sorted. Note that a sorted edge list has
the minimal memory peak, but requires the nodes number and the edges number.
from_schema: https://w3id.org/neat-ml
range: boolean
edge_list_is_correct:
name: edge_list_is_correct
description: Whether the edges file can be assumed to be correct, i.e. does not
have something wrong in it. If this parameter is passed as true on a malformed
file, the constructor will crash.
from_schema: https://w3id.org/neat-ml
range: boolean
edge_list_max_rows_number:
name: edge_list_max_rows_number
description: The maximum number of lines to be loaded from the edges file.
from_schema: https://w3id.org/neat-ml
range: integer
edge_list_comment_symbol:
name: edge_list_comment_symbol
description: The comment symbol to skip lines in the edges file. Lines starting
with this symbol will be skipped.
from_schema: https://w3id.org/neat-ml
edges_number:
name: edges_number
description: The expected number of edges. Note that this must be the EXACT number
of edges in the graph.
from_schema: https://w3id.org/neat-ml
range: integer
load_edge_list_in_parallel:
name: load_edge_list_in_parallel
description: Whether to load the edge list in parallel. Note that, if the edge
IDs indices are not given, it is NOT possible to load a sorted edge list. Similarly,
when loading in parallel, without edge IDs, the edges may not be loaded in a
deterministic order.
from_schema: https://w3id.org/neat-ml
range: boolean
verbose:
name: verbose
description: Whether to show a loading bar while reading the files. Note that,
if parallel loading is enabled, loading bars will not be showed because they
are a synchronization bottleneck.
from_schema: https://w3id.org/neat-ml
range: boolean
may_have_singletons:
name: may_have_singletons
description: Whether the graph may be expected to have singleton nodes. If it
is said that it surely DOES NOT have any, it will allow for some speedups and
lower mempry peaks.
from_schema: https://w3id.org/neat-ml
range: boolean
may_have_singleton_with_selfloops:
name: may_have_singleton_with_selfloops
description: Whether the graph may be expected to have singleton nodes with selfloops.
If it is said that it surely DOES NOT have any, it will allow for some speedups
and lower mempry peaks.
from_schema: https://w3id.org/neat-ml
range: boolean
directed:
name: directed
description: Whether to load the graph as directed or undirected.
from_schema: https://w3id.org/neat-ml
range: boolean
name:
name: name
from_schema: https://w3id.org/neat-ml
Induced
name: EnsmallenRunConfig
description: All params used by ensmallen's csv_reader
from_schema: https://w3id.org/neat-ml
attributes:
node_type_path:
name: node_type_path
description: The path to the file with the unique node type names.
from_schema: https://w3id.org/neat-ml
alias: node_type_path
owner: EnsmallenRunConfig
range: string
node_type_list_separator:
name: node_type_list_separator
description: The separator to use for the node types file.
from_schema: https://w3id.org/neat-ml
alias: node_type_list_separator
owner: EnsmallenRunConfig
range: string
node_types_column_number:
name: node_types_column_number
description: Node type column number.
from_schema: https://w3id.org/neat-ml
alias: node_types_column_number
owner: EnsmallenRunConfig
range: integer
node_types_column:
name: node_types_column
description: Node type column name.
from_schema: https://w3id.org/neat-ml
alias: node_types_column
owner: EnsmallenRunConfig
range: string
node_types_ids_column_number:
name: node_types_ids_column_number
description: Node type ID column number.
from_schema: https://w3id.org/neat-ml
alias: node_types_ids_column_number
owner: EnsmallenRunConfig
range: integer
node_types_ids_column:
name: node_types_ids_column
description: Node type ID column name.
from_schema: https://w3id.org/neat-ml
alias: node_types_ids_column
owner: EnsmallenRunConfig
range: string
node_types_number:
name: node_types_number
description: The number of the unique node types. This will be used in order to
allocate the correct size for the data structure.
from_schema: https://w3id.org/neat-ml
alias: node_types_number
owner: EnsmallenRunConfig
range: integer
numeric_node_type_ids:
name: numeric_node_type_ids
description: Whether the node type names should be loaded as numeric values, i.e.
casted from string to a numeric representation.
from_schema: https://w3id.org/neat-ml
alias: numeric_node_type_ids
owner: EnsmallenRunConfig
range: boolean
minimum_node_type_id:
name: minimum_node_type_id
description: The minimum node type ID to be used when using numeric node type
IDs.
from_schema: https://w3id.org/neat-ml
alias: minimum_node_type_id
owner: EnsmallenRunConfig
range: integer
node_type_list_header:
name: node_type_list_header
description: Whether the node type file has an header.
from_schema: https://w3id.org/neat-ml
alias: node_type_list_header
owner: EnsmallenRunConfig
range: boolean
node_type_list_support_balanced_quotes:
name: node_type_list_support_balanced_quotes
description: Whether to support balanced quotes.
from_schema: https://w3id.org/neat-ml
alias: node_type_list_support_balanced_quotes
owner: EnsmallenRunConfig
range: boolean
node_type_list_rows_to_skip:
name: node_type_list_rows_to_skip
description: The number of lines to skip in the node types file. The header is
already skipped if it has been specified that the file has an header.
from_schema: https://w3id.org/neat-ml
alias: node_type_list_rows_to_skip
owner: EnsmallenRunConfig
range: integer
node_type_list_is_correct:
name: node_type_list_is_correct
description: Whether the node types file can be assumed to be correct, i.e. does
not have something wrong in it. If this parameter is passed as true on a malformed
file, the constructor will crash.
from_schema: https://w3id.org/neat-ml
alias: node_type_list_is_correct
owner: EnsmallenRunConfig
range: boolean
node_type_list_max_rows_number:
name: node_type_list_max_rows_number
description: The maximum number of lines to be loaded from the node types file.
from_schema: https://w3id.org/neat-ml
alias: node_type_list_max_rows_number
owner: EnsmallenRunConfig
range: integer
node_type_list_comment_symbol:
name: node_type_list_comment_symbol
description: The comment symbol to skip lines in the node types file. Lines starting
with this symbol will be skipped.
from_schema: https://w3id.org/neat-ml
alias: node_type_list_comment_symbol
owner: EnsmallenRunConfig
range: string
load_node_type_list_in_parallel:
name: load_node_type_list_in_parallel
description: Whether to load the node type list in parallel. Note that when loading
in parallel, the internal order of the node type IDs may result changed across
different iterations. We are working to get this to be stable.
from_schema: https://w3id.org/neat-ml
alias: load_node_type_list_in_parallel
owner: EnsmallenRunConfig
range: boolean
node_path:
name: node_path
description: The path to the file with the unique node names.
from_schema: https://w3id.org/neat-ml
alias: node_path
owner: EnsmallenRunConfig
range: string
node_list_separator:
name: node_list_separator
description: The separator to use for the nodes file. Note that if this is not
provided, one will be automatically detected among the following - comma, semi-column,
tab and space.
from_schema: https://w3id.org/neat-ml
alias: node_list_separator
owner: EnsmallenRunConfig
range: string
node_list_header:
name: node_list_header
description: Whether the nodes file has an header.
from_schema: https://w3id.org/neat-ml
alias: node_list_header
owner: EnsmallenRunConfig
range: boolean
node_list_support_balanced_quotes:
name: node_list_support_balanced_quotes
description: Whether to support balanced quotes.
from_schema: https://w3id.org/neat-ml
alias: node_list_support_balanced_quotes
owner: EnsmallenRunConfig
range: boolean
node_list_rows_to_skip:
name: node_list_rows_to_skip
description: Number of rows to skip in the node list file.
from_schema: https://w3id.org/neat-ml
alias: node_list_rows_to_skip
owner: EnsmallenRunConfig
range: integer
node_list_is_correct:
name: node_list_is_correct
description: Whether the nodes file can be assumed to be correct, i.e. does not
have something wrong in it. If this parameter is passed as true on a malformed
file, the constructor will crash.
from_schema: https://w3id.org/neat-ml
alias: node_list_is_correct
owner: EnsmallenRunConfig
range: boolean
node_list_max_rows_number:
name: node_list_max_rows_number
description: The maximum number of lines to be loaded from the nodes file.
from_schema: https://w3id.org/neat-ml
alias: node_list_max_rows_number
owner: EnsmallenRunConfig
range: integer
node_list_comment_symbol:
name: node_list_comment_symbol
description: The comment symbol to skip lines in the nodes file. Lines starting
with this symbol will be skipped.
from_schema: https://w3id.org/neat-ml
alias: node_list_comment_symbol
owner: EnsmallenRunConfig
range: string
default_node_type:
name: default_node_type
description: The node type to be used when the node type for a given node in the
node file is None.
from_schema: https://w3id.org/neat-ml
alias: default_node_type
owner: EnsmallenRunConfig
range: string
nodes_column_number:
name: nodes_column_number
description: The number of the column of the node file from where to load the
node names.
from_schema: https://w3id.org/neat-ml
alias: nodes_column_number
owner: EnsmallenRunConfig
range: integer
nodes_column:
name: nodes_column
description: The name of the column of the node file from where to load the node
names.
from_schema: https://w3id.org/neat-ml
alias: nodes_column
owner: EnsmallenRunConfig
range: string
node_types_separator:
name: node_types_separator
description: The node types separator.
from_schema: https://w3id.org/neat-ml
alias: node_types_separator
owner: EnsmallenRunConfig
range: string
node_list_node_types_column_number:
name: node_list_node_types_column_number
description: The number of the column of the node file from where to load the
node types.
from_schema: https://w3id.org/neat-ml
alias: node_list_node_types_column_number
owner: EnsmallenRunConfig
range: integer
node_list_node_types_column:
name: node_list_node_types_column
description: The name of the column of the node file from where to load the node
types.
from_schema: https://w3id.org/neat-ml
alias: node_list_node_types_column
owner: EnsmallenRunConfig
range: string
node_ids_column:
name: node_ids_column
description: The name of the column of the node file from where to load the node
IDs.
from_schema: https://w3id.org/neat-ml
alias: node_ids_column
owner: EnsmallenRunConfig
range: string
node_ids_column_number:
name: node_ids_column_number
description: The number of the column of the node file from where to load the
node IDs
from_schema: https://w3id.org/neat-ml
alias: node_ids_column_number
owner: EnsmallenRunConfig
range: integer
nodes_number:
name: nodes_number
description: The expected number of nodes. Note that this must be the EXACT number
of nodes in the graph.
from_schema: https://w3id.org/neat-ml
alias: nodes_number
owner: EnsmallenRunConfig
range: integer
minimum_node_id:
name: minimum_node_id
description: The minimum node ID to be used, when loading the node IDs as numerical.
from_schema: https://w3id.org/neat-ml
alias: minimum_node_id
owner: EnsmallenRunConfig
range: integer
numeric_node_ids:
name: numeric_node_ids
description: Whether to load the numeric node IDs as numeric.
from_schema: https://w3id.org/neat-ml
alias: numeric_node_ids
owner: EnsmallenRunConfig
range: boolean
node_list_numeric_node_type_ids:
name: node_list_numeric_node_type_ids
description: Whether to load the node types IDs in the node file to be numeric.
from_schema: https://w3id.org/neat-ml
alias: node_list_numeric_node_type_ids
owner: EnsmallenRunConfig
range: boolean
skip_node_types_if_unavailable:
name: skip_node_types_if_unavailable
description: Whether to skip the node types without raising an error if these
are unavailable.
from_schema: https://w3id.org/neat-ml
alias: skip_node_types_if_unavailable
owner: EnsmallenRunConfig
range: boolean
load_node_list_in_parallel:
name: load_node_list_in_parallel
description: Whether to load the node list in parallel. When loading in parallel,
without node IDs, the nodes may not be loaded in a deterministic order.
from_schema: https://w3id.org/neat-ml
alias: load_node_list_in_parallel
owner: EnsmallenRunConfig
range: boolean
edge_type_path:
name: edge_type_path
description: The path to the file with the unique edge type names.
from_schema: https://w3id.org/neat-ml
alias: edge_type_path
owner: EnsmallenRunConfig
range: string
edge_types_column_number:
name: edge_types_column_number
description: The number of the column of the edge types file from where to load
the edge types.
from_schema: https://w3id.org/neat-ml
alias: edge_types_column_number
owner: EnsmallenRunConfig
range: integer
edge_types_column:
name: edge_types_column
description: The name of the column of the edge types file from where to load
the edge types.
from_schema: https://w3id.org/neat-ml
alias: edge_types_column
owner: EnsmallenRunConfig
range: string
edge_types_number:
name: edge_types_number
description: The number of the unique edge types. This will be used in order to
allocate the correct size for the data structure.
from_schema: https://w3id.org/neat-ml
alias: edge_types_number
owner: EnsmallenRunConfig
range: integer
numeric_edge_type_ids:
name: numeric_edge_type_ids
description: Whether the edge type names should be loaded as numeric values, i.e.
casted from string to a numeric representation.
from_schema: https://w3id.org/neat-ml
alias: numeric_edge_type_ids
owner: EnsmallenRunConfig
range: boolean
minimum_edge_type_id:
name: minimum_edge_type_id
description: The minimum edge type ID to be used when using numeric edge type
IDs.
from_schema: https://w3id.org/neat-ml
alias: minimum_edge_type_id
owner: EnsmallenRunConfig
range: integer
edge_type_list_separator:
name: edge_type_list_separator
description: The separator to use for the edge type list. Note that, if None is
provided, one will be attempted to be detected automatically between ';', ',',
tab or space.
from_schema: https://w3id.org/neat-ml
alias: edge_type_list_separator
owner: EnsmallenRunConfig
range: string
edge_type_list_header:
name: edge_type_list_header
description: Whether the edge type file has an header.
from_schema: https://w3id.org/neat-ml
alias: edge_type_list_header
owner: EnsmallenRunConfig
range: boolean
edge_type_list_support_balanced_quotes:
name: edge_type_list_support_balanced_quotes
description: Whether to support balanced quotes while reading the edge type list.
from_schema: https://w3id.org/neat-ml
alias: edge_type_list_support_balanced_quotes
owner: EnsmallenRunConfig
range: boolean
edge_type_list_rows_to_skip:
name: edge_type_list_rows_to_skip
description: Number of rows to skip in the edge type list file.
from_schema: https://w3id.org/neat-ml
alias: edge_type_list_rows_to_skip
owner: EnsmallenRunConfig
range: integer
edge_type_list_is_correct:
name: edge_type_list_is_correct
description: Whether the edge types file can be assumed to be correct, i.e. does
not have something wrong in it. If this parameter is passed as true on a malformed
file, the constructor will crash.
from_schema: https://w3id.org/neat-ml
alias: edge_type_list_is_correct
owner: EnsmallenRunConfig
range: boolean
edge_type_list_max_rows_number:
name: edge_type_list_max_rows_number
description: The maximum number of lines to be loaded from the edge types file.
from_schema: https://w3id.org/neat-ml
alias: edge_type_list_max_rows_number
owner: EnsmallenRunConfig
range: integer
edge_type_list_comment_symbol:
name: edge_type_list_comment_symbol
description: The comment symbol to skip lines in the edge types file. Lines starting
with this symbol will be skipped.
from_schema: https://w3id.org/neat-ml
alias: edge_type_list_comment_symbol
owner: EnsmallenRunConfig
range: string
load_edge_type_list_in_parallel:
name: load_edge_type_list_in_parallel
description: Whether to load the edge type list in parallel. When loading in parallel,
without edge type IDs, the edge types may not be loaded in a deterministic order.
from_schema: https://w3id.org/neat-ml
alias: load_edge_type_list_in_parallel
owner: EnsmallenRunConfig
range: boolean
edge_path:
name: edge_path
description: The path to the file with the edge list.
from_schema: https://w3id.org/neat-ml
alias: edge_path
owner: EnsmallenRunConfig
range: string
edge_list_separator:
name: edge_list_separator
description: The separator to use for the edge list. Note that, if None is provided,
one will be attempted to be detected automatically between ';', ',', tab or
space.
from_schema: https://w3id.org/neat-ml
alias: edge_list_separator
owner: EnsmallenRunConfig
range: string
edge_list_header:
name: edge_list_header
description: Whether the edges file has an header.
from_schema: https://w3id.org/neat-ml
alias: edge_list_header
owner: EnsmallenRunConfig
range: boolean
edge_list_support_balanced_quotes:
name: edge_list_support_balanced_quotes
description: Whether to support balanced quotes while reading the edge list.
from_schema: https://w3id.org/neat-ml
alias: edge_list_support_balanced_quotes
owner: EnsmallenRunConfig
range: boolean
edge_list_rows_to_skip:
name: edge_list_rows_to_skip
description: Number of rows to skip in the edge list file.
from_schema: https://w3id.org/neat-ml
alias: edge_list_rows_to_skip
owner: EnsmallenRunConfig
range: integer
sources_column_number:
name: sources_column_number
description: The number of the column of the edges file from where to load the
source nodes.
from_schema: https://w3id.org/neat-ml
alias: sources_column_number
owner: EnsmallenRunConfig
range: integer
sources_column:
name: sources_column
description: The name of the column of the edges file from where to load the source
nodes.
from_schema: https://w3id.org/neat-ml
alias: sources_column
owner: EnsmallenRunConfig
range: string
destinations_column_number:
name: destinations_column_number
description: The number of the column of the edges file from where to load the
destinaton nodes.
from_schema: https://w3id.org/neat-ml
alias: destinations_column_number
owner: EnsmallenRunConfig
range: integer
destinations_column:
name: destinations_column
description: The name of the column of the edges file from where to load the destinaton
nodes.
from_schema: https://w3id.org/neat-ml
alias: destinations_column
owner: EnsmallenRunConfig
range: string
edge_list_edge_types_column_number:
name: edge_list_edge_types_column_number
description: The number of the column of the edges file from where to load the
edge types.
from_schema: https://w3id.org/neat-ml
alias: edge_list_edge_types_column_number
owner: EnsmallenRunConfig
range: integer
edge_list_edge_types_column:
name: edge_list_edge_types_column
description: The name of the column of the edges file from where to load the edge
types.
from_schema: https://w3id.org/neat-ml
alias: edge_list_edge_types_column
owner: EnsmallenRunConfig
range: string
default_edge_type:
name: default_edge_type
description: The edge type to be used when the edge type for a given edge in the
edge file is None.
from_schema: https://w3id.org/neat-ml
alias: default_edge_type
owner: EnsmallenRunConfig
range: string
weights_column_number:
name: weights_column_number
description: The number of the column of the edges file from where to load the
edge weights.
from_schema: https://w3id.org/neat-ml
alias: weights_column_number
owner: EnsmallenRunConfig
range: integer
weights_column:
name: weights_column
description: The name of the column of the edges file from where to load the edge
weights.
from_schema: https://w3id.org/neat-ml
alias: weights_column
owner: EnsmallenRunConfig
range: string
default_weight:
name: default_weight
description: The edge weight to be used when the edge weight for a given edge
in the edge file is None.
from_schema: https://w3id.org/neat-ml
alias: default_weight
owner: EnsmallenRunConfig
range: float
edge_ids_column:
name: edge_ids_column
description: The name of the column of the edges file from where to load the edge
IDs.
from_schema: https://w3id.org/neat-ml
alias: edge_ids_column
owner: EnsmallenRunConfig
range: string
edge_ids_column_number:
name: edge_ids_column_number
description: The number of the column of the edges file from where to load the
edge IDs.
from_schema: https://w3id.org/neat-ml
alias: edge_ids_column_number
owner: EnsmallenRunConfig
range: integer
edge_list_numeric_edge_type_ids:
name: edge_list_numeric_edge_type_ids
description: Whether to load the edge type IDs as numeric from the edge list.
from_schema: https://w3id.org/neat-ml
alias: edge_list_numeric_edge_type_ids
owner: EnsmallenRunConfig
range: boolean
edge_list_numeric_node_ids:
name: edge_list_numeric_node_ids
description: Whether to load the edge node IDs as numeric from the edge list.
from_schema: https://w3id.org/neat-ml
alias: edge_list_numeric_node_ids
owner: EnsmallenRunConfig
range: boolean
skip_weights_if_unavailable:
name: skip_weights_if_unavailable
description: Whether to skip the weights without raising an error if these are
unavailable.
from_schema: https://w3id.org/neat-ml
alias: skip_weights_if_unavailable
owner: EnsmallenRunConfig
range: boolean
skip_edge_types_if_unavailable:
name: skip_edge_types_if_unavailable
description: Whether to skip the edge types without raising an error if these
are unavailable.
from_schema: https://w3id.org/neat-ml
alias: skip_edge_types_if_unavailable
owner: EnsmallenRunConfig
range: boolean
edge_list_is_complete:
name: edge_list_is_complete
description: Whether to consider the edge list as complete, i.e. the edges are
presented in both directions when loading an undirected graph.
from_schema: https://w3id.org/neat-ml
alias: edge_list_is_complete
owner: EnsmallenRunConfig
range: boolean
edge_list_may_contain_duplicates:
name: edge_list_may_contain_duplicates
description: Whether the edge list may contain duplicates. If the edge list surely
DOES NOT contain duplicates, a validation step may be skipped. By default, it
is assumed that the edge list may contain duplicates.
from_schema: https://w3id.org/neat-ml
alias: edge_list_may_contain_duplicates
owner: EnsmallenRunConfig
range: boolean
edge_list_is_sorted:
name: edge_list_is_sorted
description: Whether the edge list is sorted. Note that a sorted edge list has
the minimal memory peak, but requires the nodes number and the edges number.
from_schema: https://w3id.org/neat-ml
alias: edge_list_is_sorted
owner: EnsmallenRunConfig
range: boolean
edge_list_is_correct:
name: edge_list_is_correct
description: Whether the edges file can be assumed to be correct, i.e. does not
have something wrong in it. If this parameter is passed as true on a malformed
file, the constructor will crash.
from_schema: https://w3id.org/neat-ml
alias: edge_list_is_correct
owner: EnsmallenRunConfig
range: boolean
edge_list_max_rows_number:
name: edge_list_max_rows_number
description: The maximum number of lines to be loaded from the edges file.
from_schema: https://w3id.org/neat-ml
alias: edge_list_max_rows_number
owner: EnsmallenRunConfig
range: integer
edge_list_comment_symbol:
name: edge_list_comment_symbol
description: The comment symbol to skip lines in the edges file. Lines starting
with this symbol will be skipped.
from_schema: https://w3id.org/neat-ml
alias: edge_list_comment_symbol
owner: EnsmallenRunConfig
range: string
edges_number:
name: edges_number
description: The expected number of edges. Note that this must be the EXACT number
of edges in the graph.
from_schema: https://w3id.org/neat-ml
alias: edges_number
owner: EnsmallenRunConfig
range: integer
load_edge_list_in_parallel:
name: load_edge_list_in_parallel
description: Whether to load the edge list in parallel. Note that, if the edge
IDs indices are not given, it is NOT possible to load a sorted edge list. Similarly,
when loading in parallel, without edge IDs, the edges may not be loaded in a
deterministic order.
from_schema: https://w3id.org/neat-ml
alias: load_edge_list_in_parallel
owner: EnsmallenRunConfig
range: boolean
verbose:
name: verbose
description: Whether to show a loading bar while reading the files. Note that,
if parallel loading is enabled, loading bars will not be showed because they
are a synchronization bottleneck.
from_schema: https://w3id.org/neat-ml
alias: verbose
owner: EnsmallenRunConfig
range: boolean
may_have_singletons:
name: may_have_singletons
description: Whether the graph may be expected to have singleton nodes. If it
is said that it surely DOES NOT have any, it will allow for some speedups and
lower mempry peaks.
from_schema: https://w3id.org/neat-ml
alias: may_have_singletons
owner: EnsmallenRunConfig
range: boolean
may_have_singleton_with_selfloops:
name: may_have_singleton_with_selfloops
description: Whether the graph may be expected to have singleton nodes with selfloops.
If it is said that it surely DOES NOT have any, it will allow for some speedups
and lower mempry peaks.
from_schema: https://w3id.org/neat-ml
alias: may_have_singleton_with_selfloops
owner: EnsmallenRunConfig
range: boolean
directed:
name: directed
description: Whether to load the graph as directed or undirected.
from_schema: https://w3id.org/neat-ml
alias: directed
owner: EnsmallenRunConfig
range: boolean
name:
name: name
from_schema: https://w3id.org/neat-ml
alias: name
owner: EnsmallenRunConfig
range: string