Window

LEAD

Snowflake

Mengakses nilai dari baris selanjutnya dalam result set. Kebalikan dari LAG.

Tipe hasil: Same as expressionDiperbarui: 6 Jan 2026

Syntax

SQL

LEAD(expression [, offset [, default]]) OVER ([PARTITION BY partition_expr] ORDER BY order_expr)

Parameter

expressionanywajib

Kolom atau ekspresi yang nilainya akan diambil

offsetINTEGERopsional

Berapa baris ke depan (default: 1)

Default: 1

defaultanyopsional

Nilai default jika tidak ada baris selanjutnya

Contoh Penggunaan

Melihat Event Selanjutnya

SQL

1 SELECT
2   event_name,
3   event_date,
4   LEAD(event_name) OVER (ORDER BY event_date) as next_event,
5   LEAD(event_date) OVER (ORDER BY event_date) as next_event_date
6 FROM events;

Menampilkan event saat ini dan event selanjutnya.

Hasil

EVENT_NAME	EVENT_DATE	NEXT_EVENT	NEXT_EVENT_DATE
Design	2024-03-10	Development	2024-03-20
Development	2024-03-20	Testing	2024-03-25
Testing	2024-03-25	NULL	NULL

Menghitung Durasi Event

SQL

1 SELECT
2   event_name,
3   start_time,
4   LEAD(start_time) OVER (ORDER BY start_time) as next_start,
5   DATEDIFF('MINUTE', start_time,
6     LEAD(start_time) OVER (ORDER BY start_time)
7   ) as duration_minutes
8 FROM events;

Menghitung durasi setiap event (sampai event berikutnya dimulai).

Prediksi Next Action

SQL

1 SELECT
2   user_id,
3   action,
4   action_time,
5   LEAD(action, 1, 'SESSION_END') OVER (
6     PARTITION BY user_id
7     ORDER BY action_time
8   ) as next_action
9 FROM user_actions;

Melihat aksi selanjutnya per user, dengan default 'SESSION_END' jika tidak ada.

Hasil

USER_ID	ACTION	ACTION_TIME	NEXT_ACTION
U001	page_view	10:00:00	click
U001	click	10:01:00	purchase
U001	purchase	10:05:00	SESSION_END

Fungsi Terkait

LAGFIRST_VALUELAST_VALUENTH_VALUE

Kembali ke Snowflake Functions

1	SELECT
2	event_name,
3	event_date,
4	LEAD(event_name) OVER (ORDER BY event_date) as next_event,
5	LEAD(event_date) OVER (ORDER BY event_date) as next_event_date
6	FROM events;

1	SELECT
2	event_name,
3	start_time,
4	LEAD(start_time) OVER (ORDER BY start_time) as next_start,
5	DATEDIFF('MINUTE', start_time,
6	LEAD(start_time) OVER (ORDER BY start_time)
7	) as duration_minutes
8	FROM events;

1	SELECT
2	user_id,
3	action,
4	action_time,
5	LEAD(action, 1, 'SESSION_END') OVER (
6	PARTITION BY user_id
7	ORDER BY action_time
8	) as next_action
9	FROM user_actions;