streamsight.settings.NPastInteractionTimestampSplitter

class streamsight.settings.NPastInteractionTimestampSplitter(t, t_upper: int | None = None, n_seq_data: int = 1, include_all_past_data: bool = False)

Bases: TimestampSplitter

Splits with n past interactions based on a timestamp.

Splits the data into unlabeled and ground truth data based on a timestamp. Historical data contains last n_seq_data interactions before the timestamp t and the future interaction contains interactions after the timestamp t.

Attribute definition

  • past_interaction: List of unlabeled data. Interval is [0, t).

  • future_interaction: Data used for training the model. Interval is [t, t+t_upper) or [t,inf].

param t:

Timestamp to split on in seconds since epoch.

type t:

int

param t_upper:

Seconds past t. Upper bound on the timestamp of interactions. Defaults to None (infinity).

type t_upper:

int, optional

param n_seq_data:

Number of last interactions to provide as unlabeled data for model to make prediction.

type n_seq_data:

int, optional

param include_all_past_data:

If True, include all past data in the past_interaction.

type include_all_past_data:

bool, optional

return:

A 2-tuple containing the past_interaction and future_interaction matrices.

rtype:

Tuple[InteractionMatrix, InteractionMatrix]

__init__(t, t_upper: int | None = None, n_seq_data: int = 1, include_all_past_data: bool = False)

Methods

__init__(t[, t_upper, n_seq_data, ...])

split(data)

Splits data such that the following definition holds:

update_split_point(t)

Attributes

identifier

String identifier of the splitter object, contains name and parameter values.

name

The name of the splitter.
_abc_impl = <_abc._abc_data object>
property identifier

String identifier of the splitter object, contains name and parameter values.

property name

The name of the splitter.

split(data: InteractionMatrix) → Tuple[InteractionMatrix, InteractionMatrix]

Splits data such that the following definition holds:

  • past_interaction : List of unlabeled data. Interval is [0, t).

  • future_interaction : Data used for training the model. Interval is [t, t+t_upper) or [t,inf].

Parameters:

data (InteractionMatrix) – Interaction matrix to be split. Must contain timestamps.

Returns:

A 2-tuple containing the past_interaction and future_interaction matrices.

Return type:

Tuple[InteractionMatrix, InteractionMatrix]

update_split_point(t: int)