In this lesson, we'll review how to write SQL queries to retrieve and add specific data to SQL database tables.
You will be able to:
For labs in this section, we'll be working with SQL databases, which are stored as .db (or sometimes .sqlite) files. There are two methods we can use to access the data in those SQL databases. The first is to open up a sqlite database through the terminal (Mac/Linux) or Powershell(Windows).
To open a sqlite server, we can open a terminal window and type sqlite3. This will be connected to a transient, in-memory database, because we have not connected it to anything. Additional steps will be needed to save any work done to tables.
To open a database file called cats.db, we just add the name of the file as a parameter when opening sqlite by typing sqlite3 cats.db.
By default, the Python standard libary contains a package called sqlite3. We can use this package to connect to SQL databases and query or modify them as needed.
To work with a SQL database through python, the following steps are always needed:
sqlite3 librarysqlite3 library that connects to the target databaseOnce these steps are complete, we can use the cursor object to execute SQL Commands. Note that if we make any changes to the database, we must commit them first - otherwise, they'll be lost when we disconnect from the database.
The following code block contains an example of connecting to an example database, creating a table, saving changes, and disconnecting:
# Step 1: Import library
import sqlite3
# Step 2: Create a connection object that connects to the database in question
conn = sqlite3.connect('example.db')
# Step 3: Create a cursor object and use it to execute sql commands
c = conn.cursor()
# Example of using cursor to execute sql command to create a table
c.execute('''CREATE TABLE stocks
(date text, trans text, symbol text, qty real, price real)''')
# Save the changes
conn.commit()
# Disconnect from database. Any uncommited changes will be lost.
conn.close()For a full run down of how to use the sqlite library in python, see the python docs for this library.
The term "query" refers to any SQL statement that retrieves data from your database. In fact, we've already written a number of SQL queries using basic SELECT statements. We've already seen how to retrieve single units of data, or rows, with queries like these:
To select all of the rows from a cats table:
SELECT * FROM cats;To select only rows representing data meeting certain conditions:
SELECT * FROM cats WHERE name = "Maru";What if, however, we wanted to select the oldest cat? Or all of the cats that don't currently belong to an owner? Or all of the cats with short names?
Data storage isn't very useful if we can't manipulate, view, and analyze that data. Luckily for us, SQL is a powerful tool for doing just that.
In this exercise, we'll walk through executing a handful of common and handy SQL queries.
So far, we've only queried databases, as this is a much more common task for a data scientist - its more rare to be tasked with creating or altering tables. However, this is still valuable to know - try to work through the remainder of this notebook by executing these sql commands through the terminal!
In this code along, we'll be creating a cats table in a pets_database.db. So, let's navigate to our terminal and get started.
First let's create our pets_database by running the following command.
sqlite3 pets_database.dbNow that we have a database, let's create our cats table along with id, name, age, breed, and owner_id columns.
CREATE TABLE cats (
id INTEGER PRIMARY KEY,
name TEXT,
age INTEGER,
breed TEXT,
owner_id INTEGER
);Good work. Let's type ls in the terminal and see what just happened. A new file should appear called pets_database.db! This is the binary representation of the database. You can think of this like a .jpg file. It won't open up in a text editor, but it does open up in the image viewer app. It is the same way for .db files. They won't open in your editor, but they can be read by the appropriate database engine.
Let's add some cats to our cats table to make this more interesting:
sqlite> INSERT INTO cats (name, age, breed, owner_id) VALUES ("Maru", 3 , "Scottish Fold", 1);
sqlite> INSERT INTO cats (name, age, breed, owner_id) VALUES ("Hana", 1 , "Tabby", 1);
sqlite> INSERT INTO cats (name, age, breed) VALUES ("Lil\' Bub", 5, "American Shorthair");
sqlite> INSERT INTO cats (name, age, breed) VALUES ("Moe", 10, "Tabby");
sqlite> INSERT INTO cats (name, age, breed) VALUES ("Patches", 2, "Calico");Let's check out our cats table now:
sqlite> SELECT * FROM cats;This should return:
1|Maru|3|Scottish Fold|1
2|Hana|1|Tabby|1
3|Lil\' Bub|5|American Shorthair|
4|Moe|10|Tabby|
5|Patches|2|Calico|Top-Tip: You can format the output of your select statements with a few helpful options:
.headers on # output the name of each column
.mode column # now we are in column mode, enabling us to run the next two .width commands
.width auto # adjusts and normalizes column width
# or
.width NUM1, NUM2 # customize column widthRun the first two commands and then execute the above SELECT statement instead and you should see something like this:
id name age breed owner_id
---------- ---------- ---------- ------------- ----------
1 Maru 3 Scottish Fold 1
2 Hana 1 Tabby 1
3 Lil\' Bub 5 American Shor
4 Moe 10 Tabby
5 Patches 2 Calico Much better.
ORDER BYThe first query modifier we'll explore is ORDER BY. This modifier allows us to order the table rows returned by a certain SELECT statement. Here's a boilerplate SELECT statement that uses ORDER BY:
SELECT column_name FROM table_name ORDER BY column_name ASC|DESC;Let's select our cats and order them by age:
sqlite> SELECT * FROM cats ORDER BY age;This should return the following:
id name age breed owner_id
---------- ---------- ---------- ---------- ----------
2 Hana 1 Tabby 1
5 Patches 2 Calico
1 Maru 3 Scottish F 1
3 Lil\' Bub 5 American S
4 Moe 10 Tabby When using ORDER BY, the default is to order in ascending order. If we want to specify though, we can use ASC for "ascending" or DESC for "descending." Let's try to select all of our cats and sort them by age in descending order.
sqlite> SELECT * FROM cats ORDER BY age DESC;This should return
id name age breed owner_id
---------- ---------- ---------- ---------- ----------
4 Moe 10 Tabby
3 Lil\' Bub 5 American S
1 Maru 3 Scottish F 1
5 Patches 2 Calico
2 Hana 1 Tabby 1 LIMITWhat if we want the oldest cat? If we want to select extremes from a database table––for example, the employee with the highest paycheck or the patient with the most recent appointment––we can use ORDER BY in conjunction with LIMIT.
LIMIT is used to determine the number of records you want to return from a dataset. For example:
SELECT * FROM cats ORDER BY age DESC LIMIT 1;This part of the statement: SELECT * FROM cats ORDER BY age DESC returns all of the cats in order from oldest to youngest. Setting a LIMIT of 1 returns just the first, i.e. oldest, cat on the list.
Execute the above statement in your terminal and you should see:
id name age breed owner_id
---------- ---------- ---------- ---------- ----------
4 Moe 10 Tabby Let's get the two oldest cats:
SELECT * FROM cats ORDER BY age DESC LIMIT 2;Execute that statement and you should see:
id name age breed owner_id
---------- ---------- ---------- ---------- ----------
4 Moe 10 Tabby
3 Lil\' Bub 5 American S BETWEENAs we've already established, being able to sort and select specific data sets is important. Continuing on with our example, let's say we urgently need to select all of the cats whose age is between 1 and 3. To create such a query, we can use BETWEEN. Here's an boilerplate SELECT statement using BETWEEN:
SELECT column_name(s) FROM table_name WHERE column_name BETWEEN value1 AND value2;Let's try it out on our cats table:
SELECT name FROM cats WHERE age BETWEEN 1 AND 3;This should return:
Maru
Hana
PatchesLet's say the administrator of our Pets Database has found a new cat. This kitty doesn't have a name yet, but should be added to our database right away. We can add data with missing values using the NULL keyword.
Let's insert our new cat into the database. Our abandoned kitty has a breed, but no name or age as of yet:
INSERT INTO cats (name, age, breed) VALUES (NULL, NULL, "Tabby");Now, if we look at our cats data with SELECT * FROM cats;, we should see:
id name age breed owner_id
---------- ---------- ---------- ------------- ----------
1 Maru 3 Scottish Fold 1
2 Hana 1 Tabby 1
3 Lil\' Bub 5 American Shor
4 Moe 10 Tabby
5 Patches 2 Calico
6 Tabby We can even select the mysterious, nameless kitty with the following query:
SELECT * FROM cats WHERE name IS NULL;This should return the following:
id name age breed owner_id
---------- ---------- ---------- ---------- ----------
6 Tabby COUNTNow, we'll talk about a SQL aggregate function, COUNT.
SQL aggregate functions are SQL statements that retrieve minimum and maximum values from a column, sum values in a column, get the average of a column's values, or count a number of records that meet certain conditions. You can learn more about these SQL aggregators here and here.
For now, we'll just focus on COUNT. COUNT will count the number of records that meet certain condition. Here's a standard SQL query using COUNT:
"SELECT COUNT([column name]) FROM [table name] WHERE [column name] = [value]"Let's try it out and count the number of cats who have an owner_id of 1:
SELECT COUNT(owner_id) FROM cats WHERE owner_id = 1;This should return:
COUNT(owner_id)
---------------
2 GROUP BYLastly, we'll talk about the handy aggregate function GROUP BY. Like its name
suggests, it groups your results by a given column.
Let's take our table of cats
id name age breed owner_id
---------- ---------- ---------- ------------- ----------
1 Maru 3 Scottish Fold 1
2 Hana 1 Tabby 1
3 Lil\' Bub 5 American Shor
4 Moe 10 Tabby
5 Patches 2 Calico
6 Tabby Here, we can see at a glance that there are three tabby cats and
one of every other breed — but what if we had a larger database
where we couldn't just tally up the number of cats grouped by
breed? That's where — you guessed it! — GROUP BY comes in handy.
SELECT breed, COUNT(breed) FROM cats GROUP BY breed;This should return
breed COUNT(breed)
------------------ ------------
American Shorthair 1
Calico 1
Scottish Fold 1
Tabby 3 GROUP BY is a great function for aggregating results into different segments — you can even use it on multiple columns!
SELECT breed, owner_id, COUNT(breed) FROM cats GROUP BY breed, owner_id;breed owner_id COUNT(breed)
------------------ ---------- ------------
American Shorthair 1
Calico 1
Scottish Fold 1 1
Tabby 2
Tabby 1 1 SELECTWe are now familiar with this syntax:
SELECT name FROM cats;However, you may not know that this can be written like this as well:
SELECT cats.name FROM cats;Both return:
name
----------
Maru
Hana
Lil\' Bub
Moe
Patches
SQLite allows us to explicitly state the tableName.columnName we want to select. This is particularly useful when we want data from two different tables.
Imagine we have another table called dogs with a column for the dog names:
CREATE TABLE dogs (
id INTEGER PRIMARY KEY,
name TEXT
);sqlite> INSERT INTO dogs (name) VALUES ("Clifford");If we want to get the names of all the dogs and cats, we can no longer run a query with just the column name.
SELECT name FROM cats,dogs; will return Error: ambiguous column name: name.
Instead, we must explicitly follow the tableName.columnName syntax.
SELECT cats.name, dogs.name FROM cats, dogs;You may see this in the future. Don't let it trip you up.
In this lecture you reviewed how to write SQL queries to retrieve and add specific data to SQL database tables.