Windows Function - Lead() and Lag()

Author: ptyadana

Advanced SQL for Data Science - Time Series/02.Commonly used Functions for Time Series/01.Lead.sql

@@ -0,0 +1,28 @@
+/******** Commonly Used Functions for Time Series ***********/
+
+/* we will assume last digit of server_id is department id */
+CREATE VIEW time_series.vw_utilization AS(
+	SELECT *, server_id % 10 AS dept_id
+	FROM time_series.utilization
+);
+
+SELECT * FROM time_series.vw_utilization
+LIMIT 5;
+
+
+
+--------------------- LEAD() function ---------------------------
+-- LEAD() looks forwards and allows us to compare condition with the next nth row of current row.
+-- we can also put offset of how many next rows we want to get.
+
+-- next 1 row
+SELECT dept_id, server_id, cpu_utilization,
+	LEAD(cpu_utilization) OVER (PARTITION BY dept_id ORDER BY cpu_utilization DESC)
+FROM time_series.vw_utilization
+WHERE event_time BETWEEN '2019-03-05' AND '2019-03-06';
+
+-- next 3 row
+SELECT dept_id, server_id, cpu_utilization,
+	LEAD(cpu_utilization, 3) OVER (PARTITION BY dept_id ORDER BY cpu_utilization DESC)
+FROM time_series.vw_utilization
+WHERE event_time BETWEEN '2019-03-05' AND '2019-03-06';

Advanced SQL for Data Science - Time Series/02.Commonly used Functions for Time Series/02.Lag.sql

@@ -0,0 +1,17 @@
+/******** Commonly Used Functions for Time Series ***********/
+
+--------------------- LAG() function ---------------------------
+-- to reference rows relative to the currently processed rows.
+-- LAG() looks backwards and allows us to compare condition with the previous nth row of current row.
+
+SELECT dept_id, server_id, cpu_utilization,
+	LAG(cpu_utilization) OVER (PARTITION BY dept_id ORDER BY cpu_utilization DESC)
+FROM time_series.vw_utilization
+WHERE event_time BETWEEN '2019-03-05' AND '2019-03-06';
+
+
+-- with offset of 10, looking backwards to previous 10th row from the current one
+SELECT dept_id, server_id, cpu_utilization,
+	LAG(cpu_utilization, 10) OVER (PARTITION BY dept_id ORDER BY cpu_utilization DESC)
+FROM time_series.vw_utilization
+WHERE event_time BETWEEN '2019-03-05' AND '2019-03-06';