TimeWindowSeries

Summary

This class describes a TimeWindowSeries that is applied in an Aggregation within a Batch or Stream Feature View.

Time Window Series are useful for expressing a feature like "user transaction sum for every hour in the last week."

For an overview of Aggregation Windows, check out Aggregation Windows.

Description

Tecton aggregations are applied over the specified time window series using the time_window parameter. Use the TimeWindowSeries class to create an aggregation over a series of time windows as shown in the example below:

from tecton import batch_feature_view, FilteredSource, Aggregation, TimeWindowSeries
from datetime import timedelta


@batch_feature_view(
    sources=[FilteredSource(transactions)],
    mode="spark_sql",
    entities=[user],
    aggregation_interval=timedelta(days=1),
    aggregations=[
        Aggregation(
            column="value",
            function="sum",
            time_window=TimeWindowSeries(
                series_start=timedelta(days=-7),
                window_size=timedelta(days=1),
            ),
        )
    ],
)
def user_transaction_sums(transactions):
    return f"""
        SELECT user_id, timestamp, value
        FROM {transactions}
        """

Example

Consider the following example mock data:

	user_id	timestamp	value
0	user_1	2022-05-14 00:00:00	1
1	user_1	2022-05-15 00:00:00	3
2	user_1	2022-05-16 00:00:00	6
3	user_1	2022-05-17 00:00:00	11
4	user_1	2022-05-18 00:00:00	23

During Offline Retrieval, when you pass in an events dataframe to join against, the aggregation will be computed over the time window series relative to the timestamps in the input events dataframe. We give examples of how the aggregation is computed for different timestamps in the events dataframe below.

import pandas as pd
import datetime

training_events = pd.DataFrame(
    {
        "user_id": ["user_1", "user_1", "user_1", "user_1", "user_1", "user_1"],
        "timestamp": [
            datetime(2022, 5, 15),
            datetime(2022, 5, 18),
            datetime(2022, 5, 19),
            datetime(2022, 5, 20),
            datetime(2022, 5, 24),
            datetime(2022, 5, 26),
        ],
    }
)

df = user_transaction_sums.get_features_for_events(training_events).to_pandas()
display(df)

	user_id	timestamp	user_transaction_sums__amt_sum_1d_1d_series_7d_0s_1d
0	user_1	2022-05-15 00:00:00	[None, None, None, None, None, None, 1]
1	user_1	2022-05-18 00:00:00	[None, None, None, 1, 3, 6, 11]
2	user_1	2022-05-19 00:00:00	[None, None, 1, 3, 6, 11, 23]
3	user_1	2022-05-20 00:00:00	[None, 1, 3, 6, 11, 23, None]
4	user_1	2022-05-24 00:00:00	[11, 23, None, None, None, None, None]
5	user_1	2022-05-26 00:00:00	[None, None, None, None, None, None, None]

Attributes

The attributes are the same as the __init__ method parameters. See below.

Methods

__init__(...)

Parameters

Name	Required?	Description
series_start	Yes	The relative start of the series of windows, represented as a negative timedelta.
series_end	No. Defaults to 0 (i.e. "now").	The relative end of the series of windows, represented as a negative timedelta.
window_size	Yes	The size of each window in the series, represented as a positive timedelta.
step_size	No. Defaults to window_size.	The interval by which the time windows step forward in the series, represented as a positive timedelta. This is primarily useful if you want to express a series of overlapping windows. For example, if you want a series of 3 hour windows as of every hour in the last week you would set window_size=timedelta(hours=3) and step_size=timedelta(hours=1).