Databases and SQL - Library edition

Aggregation

To get an estimate of the amount of time it will take to digitize our collection, we now want to know how many books by a given publisher are in our collection, as well as calculate the total number of pages contained in the collection. We know how to select all of the page numbers from the Works table:

SELECT Pages FROM Works;

but to combine them, we must use an aggregation function such as min or max. Each of these functions takes a set of records as input, and produces a single record as output. For example, the following statements display the smallest and the largest books in our collection:

SELECT Title, min(Pages) FROM Works;

SELECT Title, max(Pages) FROM Works;

min and max are just two of the aggregation functions built into SQL. Three others are avg, count, and sum:

SELECT avg(Pages) FROM Works;

SELECT count(Title) FROM Works Where Publisher='Wiley';

SELECT sum(Pages) FROM Works Where Publisher='Wiley';

We used count(Title) here, but we could just as easily have counted ISBN or any other field in the table, or even used count(*), since the function doesn’t care about the values themselves, just how many values there are.

SQL lets us do several aggregations at once. We can, for example, find the range of book sizes published by O’Reilly:

SELECT min(Pages), max(Pages) FROM Works Where Publisher='O''Reilly';

We can also combine aggregated results with raw results, although the output might surprise you:

SELECT Title, count(*) FROM Works Where Publisher='O''Reilly';

Why does this title appear rather than any other published by O’Reilly? The answer is that when it has to aggregate a field, but isn’t told how to, the database manager chooses an actual value from the input set. It might use the first one processed, the last one, or something else entirely.

Another important fact is that when there are no values to aggregate — for example, where there are no rows satisfying the WHERE clause — aggregation’s result is “don’t know” rather than zero or some other arbitrary value:

SELECT min(Pages), max(Pages) FROM Works Where Publisher='Hachette';

One final important feature of aggregation functions is that they are inconsistent with the rest of SQL in a very useful way. If we add two values, and one of them is null, the result is null. By extension, if we use sum to add all the values in a set, and any of those values are null, the result should also be null. It’s much more useful, though, for aggregation functions to ignore null values and only combine those that are non-null. This behavior lets us write our queries as:

SELECT Title, min(Pages) FROM Works;

instead of always having to filter explicitly:

SELECT Title, min(Pages) FROM Works WHERE Pages IS NOT NULL;

Aggregating all records at once doesn’t always make sense. For example, if we want to get a breakdown of how many books are in the collection for each publisher, we cannot write:

SELECT Publisher, count(Title) FROM Works;

because the database manager selects a single arbitrary publisher name rather than aggregating results separately for each publisher. Since in our small example there are only ten different publishers, we could write ten queries of the form:

SELECT Publisher, count(Title) FROM Works WHERE Publisher='Faber & Faber';

but this would be tedious, and it wouldn’t be practical against a real-life database containing many thousands different publishers.

What we need to do is tell the database manager to aggregate the number of titles for each publisher separately using a GROUP BY clause:

SELECT Publisher, count(Title) FROM Works GROUP BY Publisher;

GROUP BY does exactly what its name implies: groups all the records with the same value for the specified field together so that aggregation can process each batch separately. Since all the records in each batch have the same value for Publisher, it no longer matters that the database manager is picking an arbitrary one to display alongside the count of Title values.

Just as we can sort by multiple criteria at once, we can also group by multiple criteria. To get a yearly breakdown of the number of titles published by each publisher, we just add another field to the GROUP BY clause:

SELECT Publisher, Date, count(Title) FROM Works GROUP BY Publisher, Date;

Note that we have added Date to the list of fields displayed, since the results wouldn’t make much sense otherwise.

Let’s wrap up our collection analysis query:

SELECT   Publisher, Date, count(Title), sum(Pages) FROM Works 
WHERE    Pages IS NOT NULL 
GROUP BY Publisher, Date 
ORDER BY Publisher ASC, Date DESC;

Looking more closely, this query:

1. selected records from the Works table where the Pages field was not null;

2. grouped those records into subsets so that the Publisherand Date values in each subset were the same;

3. ordered those subsets first by Publisher, and then within each sub-group by `Date; and

4. counted the number of records in each subset, calculated the sum of Pages in each, and chose a Publisher and Date value from each to display (it doesn’t matter which ones, since they’re all equal in each subset)

O'Reilly, Wiley, Faber & Faber, Belknap Press, ...