Before you keep reading...

Runestone Academy can only continue if we get support from individuals like you. As a student you are well aware of the high cost of textbooks. Our mission is to provide great books to you for free, but we ask that you consider a $10 donation, more if you can or less if $10 is a burden.

Support Runestone Academy Today

Before you keep reading...

Making great stuff takes time and $$. If you appreciate the book you are reading now and want to keep quality materials free for other students please consider a donation to Runestone Academy. We ask that you consider a $10 donation, but if you can give more thats great, if $10 is too much for your budget we would be happy with whatever you can afford as a show of support.

Support Runestone Academy Today

11.5. Aggregation or Group By

One very powerful feature of SQL is that it allows us to create summary information by grouping rows together. For example, we could ask ourselves how many bike trips were taken for each subscriber type, and which subscriber type has the most bike trips.

SELECT member_type, COUNT(*) FROM trip_data GROUP BY member_type ORDER BY COUNT(*) DESC LIMIT 10

GROUP BY member_type takes all the rows with a given member_type and produces a single row in the result. This means that we need to tell SQL how we want to combine the other columns’ values into a single row. The above example uses COUNT(*) which reports the number of rows that were combined.

Aggregating the values for member_type is not hard, since they are all the same, SQL just gives us a single copy of the publisher name. Other columns, we need to either ignore (causing them to be omitted from the output) or specify a way to aggregate them.

We must specify an aggregate function for any column that we SELECT in our query (except the column that we’re grouping by) in order for the command to succeed. If we don’t specify a way to aggregate the value most database servers will complain. However, SQLITE does not. SQLite lets you do silly things without giving you an error. For example, the following query will work, but you have no idea what the results actually mean.

SELECT duration, count(*) FROM trip_data GROUP BY member_type ORDER BY COUNT(*) DESC

Here you have grouped by member_type, but without member_type in the select clause you have no idea which rows correspond to which member type. That is why most databases will flag this as a error. Furthermore the duration field may be the first duration in the group or maybe the last duration in the group or possibly in between, But it’s not defined. The best practices for writing group by queries that work well across database systems are as follows:

• Always include the GROUP BY column(s) in your SELECT clause.

• If you include a column that is not in the GROUP BY clause in your SELECT clause you must do some form of aggregation on the values in that column. For example, min, max, mean, count, etc.

Let’s go back briefly to the first query in the Aggregation section. The top result was the count of bike trips for member_type Member:

member_type	COUNT(*)
Member	979814

If you’d like to get a more granular break down of the count, you may specify multiple columns to aggregate within the GROUP BY clause, for example: further breakdown the aggregate count by the start station IDs:

SELECT member_type, start_station, count(*) FROM trip_data WHERE member_type = 'Casual' GROUP BY member_type, start_station ORDER BY COUNT(*) DESC LIMIT 20

Great! Now that you’re familiar with how to aggregate data using SQL query by using COUNT() as your aggregation function, let’ take a look at other aggregation functions.

There are many such functions. Some common ones include:

• SUM: To add the values together

• AVG: To compute the mean of the values

• MIN or MAX: To compute the minimum and maximum respectively

So we could for example compute the total number of minutes of all bike trips for all subscriber types

SELECT member_type, SUM(duration) FROM trip_data GROUP BY member_type LIMIT 10

11.5.1. Practice Exercises

SELECT start_station, AVG(duration) FROM trip_data GROUP BY start_station ORDER BY AVG(duration) DESC LIMIT 10

2. select bike_number, count(*) from trip_data group by bike_number order by count(*) desc`

3. select member_type, count(*) from trip_data group by member_type;

4. select start_station, count(*) from trip_data where start_station = end_station group by start_station order by count(*) desc

Compute the average duration of bike trips for each starting station id and list the results in order of highest average to lowest average for the 10 stations with the highest average. What is the highest average duration?

==== assert 0,1 == 40669.5

What is the bike_number and count of the bike with the most rides?

==== assert 0,0 == W00893 assert 0,1 == 1584

How many total rides by Members and Casual users?

==== assert 0,0 == Casual assert 0,1 == 246949 assert 1,0 == Member assert 1,1 == 979814

What is the station that has the most rides that start and end at the same station? How many rides started there?

==== assert 0,0 == 31217 assert 0,1 == 3135

You have attempted of activities on this page