Lab 11 Solutions

Solution Files

Attendance

If you are in a regular 61A lab, your TA will come around and check you in. You need to submit the lab problems in addition to attending to get credit for lab. If you are in the mega lab, you only need to submit the lab problems to get credit.

If you miss lab for a good reason (such as sickness or a scheduling conflict) or you don't get checked in for some reason, just fill out this form within two weeks to receive attendance credit.

Required Questions

SQL

A SELECT statement describes an output table based on input rows. To write one:

  1. Describe the input rows using FROM and WHERE clauses.
  2. Group those rows and determine which groups should appear as output rows using GROUP BY and HAVING clauses.
  3. Format and order the output rows and columns using SELECT and ORDER BY clauses.

SELECT (Step 3) FROM (Step 1) WHERE (Step 1) GROUP BY (Step 2) HAVING (Step 2) ORDER BY (Step 3);

Step 1 may involve joining tables (using commas) to form input rows that consist of two or more rows from existing tables.

The WHERE, GROUP BY, HAVING, and ORDER BY clauses are optional.

Consult the drop-down for a refresher on SQL. It's okay to skip directly to the questions and refer back here should you get stuck.

SQL Basics

Example Table

Here's a table called big_game used in the examples below, which records the scores for the Big Game each year. This table has three columns: berkeley, stanford, and year.

CREATE TABLE big_game (
    berkeley INTEGER, stanford INTEGER, year INTEGER);

INSERT INTO big_game (berkeley, stanford, year) VALUES
    (30, 7, 2002),
    (28, 16, 2003),
    (17, 38, 2014);

You can view it in sqlite like this:

sqlite> .mode column
sqlite> SELECT * FROM big_game;
berkeley  stanford  year
--------  --------  ----
30        7         2002
28        16        2003
17        38        2014

If the .mode column command doesn't work in your version of sqlite, that's ok. Just ignore it.

Selecting From Tables

Typically, we will create a new table from existing tables using a SELECT statement:

SELECT [columns] FROM [tables] WHERE [condition] ORDER BY [columns] LIMIT [limit];

Let's break down this statement:

  • SELECT [columns] tells SQL that we want to include the given columns in our output table; [columns] is a comma-separated list of column names, and * can be used to select all columns
  • FROM [table] tells SQL that the columns we want to select are from the given table
  • WHERE [condition] filters the output table by only including rows whose values satisfy the given [condition], a boolean expression
  • ORDER BY [columns] orders the rows in the output table by the given comma-separated list of columns; by default, values are sorted in ascending order (ASC), but you can use DESC to sort in descending order
  • LIMIT [limit] limits the number of rows in the output table by the integer [limit]

Here are some examples:

Select all of Berkeley's scores from the big_game table, but only include scores from years past 2002:

sqlite> SELECT berkeley FROM big_game WHERE year > 2002;
28
17

Select the scores for both schools in years that Berkeley won:

sqlite> SELECT berkeley, stanford FROM big_game WHERE berkeley > stanford;
30|7
28|16

Select the years that Stanford scored more than 15 points:

sqlite> SELECT year FROM big_game WHERE stanford > 15;
2003
2014

SQL operators

Expressions in the SELECT, WHERE, and ORDER BY clauses can contain one or more of the following operators:

  • comparison operators: =, >, <, <=, >=, <> or != ("not equal")
  • boolean operators: AND, OR
  • arithmetic operators: +, -, *, /
  • concatenation operator: ||

Output the ratio of Berkeley's score to Stanford's score each year:

sqlite> select berkeley * 1.0 / stanford from big_game;
0.447368421052632
1.75
4.28571428571429

Output the sum of scores in years where both teams scored over 10 points:

sqlite> select berkeley + stanford from big_game where berkeley > 10 and stanford > 10;
55
44

Output a table with a single column and single row containing the value "hello world":

sqlite> SELECT "hello" || " " || "world";
hello world

Joins

To select data from multiple tables, we can use joins. There are many types of joins, but the only one we'll worry about is the inner join. To perform an inner join on two on more tables, simply list them all out in the FROM clause of a SELECT statement:

SELECT [columns] FROM [table1], [table2], ... WHERE [condition] ORDER BY [columns] LIMIT [limit];

We can select from multiple different tables or from the same table multiple times.

Let's say we have the following table that contains the names of head football coaches at Cal since 2002:

CREATE TABLE coaches (
    name TEXT,
    start INTEGER,
    end INTEGER
);

INSERT INTO coaches (name, start, end)
VALUES
    ('Jeff Tedford', 2002, 2012),
    ('Sonny Dykes', 2013, 2016),
    ('Justin Wilcox', 2017, 2025);

When we join two or more tables, the default output is a cartesian product. For example, if we joined big_game with coaches, we'd get the following:

sqlite> SELECT * FROM big_game JOIN coaches;
berkeley  stanford  year  name           start  end
--------  --------  ----  -------------  -----  ----
30        7         2002  Jeff Tedford   2002   2012
30        7         2002  Sonny Dykes    2013   2016
30        7         2002  Justin Wilcox  2017   2025
28        16        2003  Jeff Tedford   2002   2012
28        16        2003  Sonny Dykes    2013   2016
28        16        2003  Justin Wilcox  2017   2025
17        38        2014  Jeff Tedford   2002   2012
17        38        2014  Sonny Dykes    2013   2016
17        38        2014  Justin Wilcox  2017   2025

If we want to match up each game with the coach that season, we'd have to compare columns from the two tables in the WHERE clause:

sqlite> SELECT * FROM big_game JOIN coaches ON year >= start AND year <= end
   ...> ;
berkeley  stanford  year  name          start  end
--------  --------  ----  ------------  -----  ----
30        7         2002  Jeff Tedford  2002   2012
28        16        2003  Jeff Tedford  2002   2012
17        38        2014  Sonny Dykes   2013   2016

In the query above, none of the column names are ambiguous. For example, it is clear that the start column comes from the coaches table because there isn't a column in the big_game table with that name. However, if a column name exists in more than one of the tables being joined, or if we join a table with itself, we must disambiguate the column names using dot notation and aliases.

For examples, let's find out what the score difference is for each team between a game in big_game and any previous games. Since each row in this table represents one game, in order to compare two games we must join big_game with itself:

sqlite> SELECT b.Berkeley - a.Berkeley, b.Stanford - a.Stanford, a.Year, b.Year
   ...>   FROM big_game AS a, big_game AS b WHERE a.Year < b.Year;
b.Berkeley - a.Berkeley  b.Stanford - a.Stanford  year  year
-----------------------  -----------------------  ----  ----
-2                       9                        2002  2003
-13                      31                       2002  2014
-11                      22                       2003  2014

In the query above, we give the alias a to the first big_game table and the alias b to the second big_game table. We can then reference columns from each table using dot notation with the aliases, e.g. a.Berkeley, a.Stanford, and a.Year to select from the first table.

SQL Aggregation

Here's another example table, this time about flights:

CREATE TABLE flights (
    departure TEXT, arrival TEXT, price INTEGER);

INSERT INTO flights (departure, arrival, price) VALUES
    ('SFO', 'LAX', 97),
    ('SFO', 'AUH', 848),
    ('LAX', 'SLC', 115),
    ('SFO', 'PDX', 192),
    ('AUH', 'SEA', 932),
    ('SLC', 'PDX', 79),
    ('SFO', 'LAS', 40),
    ('SLC', 'LAX', 117),
    ('SEA', 'PDX', 32),
    ('SLC', 'SEA', 42),
    ('SFO', 'SLC', 97),
    ('LAS', 'SLC', 50),
    ('LAX', 'PDX', 89);

Applying an aggregate function such as MAX(column) combines the values from multiple rows into an output row.

By default, we combine the values of all rows in the table. For example, if we wanted to count the number of rows in our flights table, we could use:

sqlite> SELECT COUNT(*) from FLIGHTS;
13

What if we wanted to group together the values in similar rows and perform the aggregation operations within those groups? We use a GROUP BY clause.

Here's another example. For each unique departure, collect all of the rows having the same departure airport into a group. Then, select the price column and apply the MIN aggregation to recover the price of the cheapest departure from that group. The end result is a table of departure airports and the cheapest departing flight.

sqlite> SELECT departure, MIN(price) FROM flights GROUP BY departure;
departure  MIN(price)
---------  ----------
AUH        932
LAS        50
LAX        89
SEA        32
SFO        40
SLC        42

Just like how we can filter out rows with WHERE, we can also filter out groups with HAVING. Typically, a HAVING clause should use an aggregation function. Suppose we want to see all airports with at least two departures:

sqlite> SELECT departure FROM flights GROUP BY departure HAVING COUNT(*) >= 2;
departure
---------
LAX
SFO
SLC

Note that the COUNT(*) aggregate just counts the number of rows in each group. Say we want to count the number of distinct airports instead. Then, we could use the following query:

sqlite> SELECT COUNT(DISTINCT departure) AS destinations FROM flights;
destinations
------------
6

This enumerates all the different departure airports available in our flights table (in this case: SFO, LAX, AUH, SLC, SEA, and LAS).

Usage

First, check that a file named sqlite_shell.py exists alongside the assignment files. If you don't see it, or if you encounter problems with it, scroll down to the Troubleshooting section to see how to download an official precompiled SQLite binary before proceeding.

You can start an interactive SQLite session in your Terminal or Git Bash with the following command:

python3 sqlite_shell.py

While the interpreter is running, you can type .help to see some of the commands you can run.

To exit out of the SQLite interpreter, type .exit or .quit or press Ctrl-C. Remember that if you see ...> after pressing enter, you probably forgot a ;.

You can also run all the statements in a .sql file by doing the following: (Here we're using the lab11.sql file as an example.)

  1. Runs your code and then exits SQLite immediately afterwards.

    python3 sqlite_shell.py < lab11.sql
  2. Runs your code and then opens an interactive SQLite session, which is similar to running Python code with the interactive -i flag.

    python3 sqlite_shell.py --init lab11.sql

Visualizing SQL

The CS61A SQL Web Interpreter is a great tool for visualizing and debugging SQL statements!

To get started, visit code.cs61a.org and hit Start SQL interpreter on the launch screen.

Most tables used in assignments are already available for use, so let's try to execute a SELECT statement:

In addition to displaying a visual representation of the output table, the "Step-by-step" button lets us step through the SQL execution and visualize every transformation that takes place. For our example, clicking on the next arrow will produce the following visuals, demonstrating exactly how SQL is grouping our rows to form the final output!


IMDb

The IMDb 1000 dataset contains the 1000 highest rated popular movies (top 0.2% of votes) from the Internet Movie Database.

The tables in this dataset are quite large, so it's often useful to use LIMIT to look at just a couple of rows in one of the tables. To look at a few example rows from the titles table, which contains information about each movie:

    % sqlite3
    sqlite> .read imdb1000.sql
    sqlite> .tables
    crew        names       principals  ratings     titles

The command beginning with .read reads in the SQL commands in imdb1000.sql, which create new tables with the movie dataset. The .tables command lists out all of the tables.

Here are the tables and columns you'll need for this lab. The .mode column command improves the formatting, but it doesn't work in the sqlite_shell.py and it's optional.

sqlite> .mode column
sqlite> SELECT tconst, title, year, runtime FROM titles LIMIT 3;
tconst     title                            year  runtime
---------  -------------------------------  ----  -------
tt0012349  The Kid                          1921  68
tt0013442  Nosferatu: A Symphony of Horror  1922  94
tt0015864  The Gold Rush                    1925  95

sqlite> SELECT tconst, ordering, nconst, character FROM principals LIMIT 3;
tconst     ordering  nconst     character
---------  --------  ---------  ---------
tt0012349  1         nm0000122  A Tramp
tt0012349  2         nm0701012  The Woman
tt0012349  3         nm0001067  The Child

sqlite> SELECT nconst, name, birth, death FROM names LIMIT 3;
nconst     name            birth  death
---------  --------------  -----  -----
nm0000002  Lauren Bacall   1924   2014
nm0000004  John Belushi    1949   1982
nm0000005  Ingmar Bergman  1918   2007

tconst is a unique identifier for each movie (the t is for title), and nconst is a unique identifier for each person. These identifiers are useful for joining rows from different tables, e.g., to determine the rating for a particular movie, we would find the movie's nconst in the titles table, and then lookup that nconst in the ratings table.

Q1: Newest Movies

Create a table newest that contains the 10 newest movies in the dataset, and that has two columns:

  • title: The name of a movie
  • year: The year the movie was made

If you'd like, see if you can guess which movies will be in the results.

  SELECT ____, ____
  FROM ____
  ORDER BY ____
  LIMIT ____;
  1. Use FROM to specify which table to read the data from.
  2. Use ORDER BY to sort by which movies are the newest.
  3. Use LIMIT to select only the 10 newest movies.
  4. Use SELECT to put the movie title and year in the output.
CREATE table newest AS
SELECT title, year FROM titles ORDER BY year DESC LIMIT 10;
Use Ok to test your code:

python3 ok -q newest

Q2: Movies with Dogs

Create a dog_movies table that includes one row for each movie character that includes the word "dog", and that has two columns:

  • title (text): The name of a movie
  • character (text): The name of the dog character

A single movie may appear multiple times if it has multiple dog characters.

SELECT _____, _____
FROM _____ JOIN _____ ON _____
WHERE ____ LIKE "%dog%";
  1. Use FROM, JOIN, and ON to combine the information in the titles and principals tables.
  2. Use WHERE to select only characters that include "dog".
  3. Use SELECT to put the movie title and character in the output.
CREATE table dog_movies AS 
SELECT title, character FROM titles JOIN principals ON titles.tconst = principals.tconst WHERE character LIKE "%dog%";
Use Ok to test your code:

python3 ok -q dog_movies

Q3: Leads of Leads

Create a leads table that has two columns:

  • name (text): The name of an actor
  • lead_roles (integer): The number of movies that the actor has been the lead in

You can find lead roles by looking in the principals table for rows with ordering set to 1. Only select actors who have been the lead in more than 10 movies.

  SELECT ____, ____ AS lead_roles
  FROM ____ JOIN ____ ON ____
  WHERE ____
  GROUP BY names.nconst
  HAVING count(*) > 10;
  1. Use FROM and ON to combine the information in the principals and names tables.
  2. Use WHERE to select only lead roles.
  3. Use GROUP BY to create a single row for each person.
  4. Use HAVING to output only people who had more than 10 lead roles.
  5. Use SELECT to put the person's name and the count of their roles in the output.
CREATE table leads AS 
SELECT name, count(*) FROM names JOIN principals ON names.nconst=principals.nconst WHERE ordering=1 GROUP BY names.nconst HAVING count(*) > 10;
Use Ok to test your code:

python3 ok -q leads

Q4: Long Movies

Create a long_movies table that contains the number of movies in each decade that are over 3 hours long. long_movies has two columns:

  • decade (strings): The decade, e.g., 1920s
  • count (numbers): The number of movies in that decade that are over 3 hours (180 minutes) long

For the IMDB 1000 dataset, the result will be:

+--------+-------+
| decade | count |
+--------+-------+
| 1930s  | 1     |
| 1950s  | 2     |
| 1960s  | 2     |
| 1970s  | 3     |
| 1980s  | 2     |
| 1990s  | 7     |
| 2000s  | 4     |
| 2010s  | 3     |
| 2020s  | 4     |
+--------+-------+
When adding numbers and including the result in a string, put parentheses around the arithmetic, e.g.,
sqlite> SELECT (192 * 10) || "s";
1920s
SELECT ____ AS decade, ____ AS count
FROM ____
WHERE ____
GROUP BY ____;
  1. Use FROM and WHERE to choose only the movies with runtime greater than 180 minutes.
  2. Use GROUP BY to group together all of the movies for one decade.
  3. Use SELECT to create a string like 1920s for each decade, and to select the count of movies for each decade.
CREATE table long_movies AS 
SELECT ((year / 10) * 10) || "s" AS decade, COUNT(*) AS count FROM titles WHERE runtime>180 GROUP BY year / 10;

Use Ok to test your code:

python3 ok -q long_movies

Check Your Score Locally

You can locally check your score on each question of this assignment by running

python3 ok --score

This does NOT submit the assignment! When you are satisfied with your score, submit the assignment to Pensieve to receive credit for it.

Submit Assignment

Submit this assignment by uploading any files you've edited to the appropriate Pensieve assignment. Lab 00 has detailed instructions.

Correctly completing all questions is worth one point. If you are in the regular lab, you will need your attendance from your TA to receive that one point. Please ensure your TA has taken your attendance before leaving.