list
List feature types.
Classes:
| Name | Description |
|---|---|
ListFeature |
Feature representing a collection of sub-features without temporal ordering. |
ListFeature
#
Feature representing a collection of sub-features without temporal ordering.
A list feature encodes a variable-length sequence of feature instances that share
the same feature type, defined by items. In contrast to TimeSeriesFeature,
a list does not imply any temporal or sequential relationship between elements;
it simply groups multiple instances of the same feature together.
This structure is suitable for representing sets or collections of homogeneous entities, such as a list of detected objects, word embeddings, or samples within a batch.
Parameters:
| Name | Type | Description | Default |
|---|---|---|---|
|
ItemType
|
Feature type of each element in the list. All elements must share the same underlying feature definition. |
required |
Notes
- The order of elements in the list is not semantically meaningful.
- This feature type is close to a
TimeSeriesFeaturewithout temporal context. It can be used for unordered sequences or collections of arbitrary length. - Current limitation: Xpdeep does not yet support variable-length lists. Please pad all lists to a uniform length before using this feature.
Methods:
| Name | Description |
|---|---|
to_model |
Convert to a |
from_model |
Instantiate from a json response. |
Attributes:
| Name | Type | Description |
|---|---|---|
items |
ItemType
|
|
name |
Literal['LIST']
|
|
items: ItemType
#
name: Literal['LIST'] = 'LIST'
#
to_model() -> ListFeatureTypeInput
#
Convert to a ListFeatureTypeInput instance.
Returns:
| Type | Description |
|---|---|
ListFeatureTypeInput
|
Model-level representation of the list feature, including its serialized item feature type. |
Source code in src/xpdeep/dataset/feature/feature_types/list.py
from_model(feature_type_input: ListFeatureTypeInput) -> ListFeature[ItemType]
#
Instantiate from a json response.