Add group_by and join exercises

Author: jfarmer

group_by.sql

@@ -0,0 +1,80 @@
+-- We can use "AS" to relabel a field inside a particular query.
+-- This not only makes the report look nicer by displaying a more useful
+-- header name, but it makes it easier to reference a field elsewhere
+-- in the query.
+
+-- "invoices" table
+
+-- A list of the top 5 countries by number of invoices
+SELECT billing_country, COUNT(*) AS invoice_count
+FROM invoices
+GROUP BY billing_country
+ORDER BY invoice_count DESC
+LIMIT 5;
+
+-- A list of the top 8 countries by gross/total invoice size
+SELECT billing_country, SUM(total) AS gross_sales
+FROM invoices
+GROUP BY billing_country
+ORDER BY gross_sales DESC
+LIMIT 8;
+
+-- A list of the top 10 countries by average invoice size
+SELECT billing_country, AVG(total) AS avg_invoice_size
+FROM invoices
+GROUP BY billing_country
+ORDER BY avg_invoice_size DESC
+LIMIT 10;
+
+-- Sales volume and receipts by year
+-- See: http://www.sqlite.org/lang_datefunc.html
+-- "STRFTIME" means "string format time" and is used to format the output of
+-- dates, times, and timestamps.
+-- See: http://cl.ly/image/0C1m372r233t
+SELECT STRFTIME('%Y', invoice_date) AS year,
+       COUNT(*) AS invoice_count,
+       SUM(total) AS invoice_total
+FROM invoices
+GROUP BY year
+ORDER BY year ASC;
+
+-- Sales volume and receipts by year and month
+-- You can group by multiple fields
+-- You can order by multiple fields (here, year first then month)
+-- See: http://cl.ly/image/2e2H3w052H2o
+
+SELECT STRFTIME('%Y', invoice_date) AS year,
+       STRFTIME('%m', invoice_date) AS month,
+       COUNT(*) AS invoice_count,
+       SUM(total) AS invoice_total
+FROM invoices
+GROUP BY year, month
+ORDER BY year ASC, month ASC;
+
+-- A list of the top 5 US states by number of invoices
+-- Hint: You'll need to filter the results with WHERE billing_country = 'USA'
+
+-- A list of the top 5 US states by gross sales
+
+-- A list of the top 5 US states by average invoice size
+
+-- A list of the top 10 US cities by number of invoices
+
+-- A list of the top 10 US cities by gross sales
+
+-- A list of the top 10 US cities by average invoice size
+
+-- A list of the top 3 cities in California by number of invoices
+-- Hint: You'll need a WHERE clause filtering by both billing_country and billing_state
+
+-- A list of the top 3 cities in California by gross sales
+
+-- A list of the top 3 cities in California by average invoice size
+
+
+-- "customers" table
+-- Remember: run ".schema customers" to see what fields (columns) the customers table contains.
+
+-- A list of the top 3 countries by total number of customers
+
+-- A list of the top 7 cities (anywhere) by total number of customers

join.sql

@@ -0,0 +1,82 @@
+-- Joins (via the JOIN clause) are how we query data contained in multiple tables.
+-- There are two main types of relationships:
+--   1. One-to-many
+--   2. Many-to-many
+
+-- For example, a customer has many invoices, but each invoice belongs to one
+-- and only one customer.  This is a one-to-many relationship.
+
+-- Think of some one-to-many relationships in every day life, e.g.,
+--
+--   1. A car can get many parking tickets, but a parking ticket belongs to
+--      one and only one car.
+--   2. Every state in the US has many cities, but every city in the US is in
+--      one and only one state.
+--   3. Customers can have many orders, but an order belongs to a single customer
+--   4. Customers can have multiple credit cards, but every credit card belong
+--      to a single customer.
+--   5. Think of some of your own.
+
+-- An example of a many-to-many relationship might be events and people.
+-- I can attend many events and each event has many attendees.
+-- In our toy database, we're only dealing with one-to-many relationships.
+
+-- The 5 customers who most recently purchased something
+SELECT customers.id, customers.first_name,
+       customers.last_name, customers.email,
+       invoices.invoice_date, invoices.total
+FROM customers
+JOIN invoices
+  ON (invoices.customer_id = customers.id)
+ORDER BY invoices.invoice_date DESC
+LIMIT 5;
+
+-- The top 10 customers by total number of orders
+SELECT customers.id, customers.first_name,
+       customers.last_name, customers.email,
+       COUNT(*) AS order_count
+FROM customers
+JOIN invoices
+  ON (invoices.customer_id = customers.id)
+GROUP BY customers.id
+ORDER BY order_count DESC
+LIMIT 10;
+
+-- The top 10 customers by gross sales
+SELECT customers.id, customers.first_name,
+       customers.last_name, customers.email,
+       SUM(total) AS gross_sales
+FROM customers
+JOIN invoices
+  ON (invoices.customer_id = customers.id)
+GROUP BY customers.id
+ORDER BY gross_sales DESC
+LIMIT 10;
+
+-- The top 5 genres by number of tracks (fill in the blanks)
+SELECT genres.id, genres.name, COUNT(*) AS track_count
+FROM genres
+JOIN tracks
+  ON (tracks.genre_id = genres.id)
+GROUP BY genres.id
+ORDER BY track_count DESC
+LIMIT 5;
+
+-- The top 5 genres by total track length (in milliseconds)
+
+-- The top 5 genres by average track length (in milliseconds)
+
+-- The top 5 albums by total track length
+-- Hint: you'll need to JOIN the albums table and the tracks table
+-- Hint: the tracks table has an album_id field
+
+-- The top 5 albums by average track length
+
+-- The top 5 albums by total album price
+-- Hint: the "tracks" table has a unit_price field, so the "price" of an album
+--       is the sum of its tracks' unit_price fields.
+
+-- The 10 albums with the longest play-time
+
+-- The 10 highest-selling tracks of all time
+-- Hint: you'll need to join the tracks table and the invoice_lines table