indexing

Author: ptyadana

Advanced SQL for Data Science - Time Series/01.Modeling Time Series Data - Indexing.sql

@@ -0,0 +1,80 @@
+/*********** Modeling Time Series Data *****************/
+
+
+------------- Load Time Series Data --------------------
+COPY time_series.location_temp(event_time, location_id, temp_celcius)
+FROM 'C:/temp-sql/location_temp.txt' DELIMITER ',';
+
+
+/* quick check the data */
+SELECT * FROM time_series.location_temp
+ORDER BY event_time
+LIMIT 5;
+
+---------------- Indexing Data ----------------------------
+
+------ Indexing on Location -----------------
+/* check avg temp of each location*/
+EXPLAIN SELECT location_id, AVG(temp_celcius)
+FROM time_series.location_temp
+GROUP BY location_id;
+
+
+/* create index and check the query execution time again */
+CREATE INDEX idx_loc_location_temp
+ON time_series.location_temp(location_id);
+
+/* we can see that it is still not using index , so we will put additional where clause to narrow down */
+EXPLAIN ANALYZE SELECT location_id, AVG(temp_celcius)
+FROM time_series.location_temp
+GROUP BY location_id;
+
+
+/* now index got used (bitmap  index) */
+EXPLAIN ANALYZE SELECT location_id, AVG(temp_celcius)
+FROM time_series.location_temp
+WHERE location_id = '2'
+GROUP BY location_id;
+
+/* let's drop index and see how the impact looks like now */
+DROP INDEX time_series.idx_loc_location_temp;
+
+
+EXPLAIN ANALYZE SELECT location_id, AVG(temp_celcius)
+FROM time_series.location_temp
+WHERE location_id = '2'
+GROUP BY location_id;
+
+
+------ Indexing on Event_Time and Location ID -----------------
+
+-- 34090
+EXPLAIN SELECT location_id, AVG(temp_celcius)
+FROM time_series.location_temp
+WHERE event_time BETWEEN '2019-03-05' AND '2019-03-06'
+GROUP BY location_id;
+
+/* create index */
+CREATE INDEX idx_even_time_location_temp
+ON time_series.location_temp(event_time, location_id);
+
+/* check again and it is still not using the index */
+EXPLAIN SELECT location_id, AVG(temp_celcius)
+FROM time_series.location_temp
+WHERE event_time BETWEEN '2019-03-05' AND '2019-03-06'
+GROUP BY location_id;
+
+
+/* now let's make more selective, adding hour, min , sec */
+EXPLAIN SELECT location_id, AVG(temp_celcius)
+FROM time_series.location_temp
+WHERE event_time BETWEEN '2019-03-05 00:00:00' AND '2019-03-05 00:20:00'
+GROUP BY location_id;
+
+
+/* 
+So in summary, Indexes are very helpful in some cases, especially when we are working with limited subset of data.
+But they don't always help when we are working with larget time span. 
+*/
+
+