Starting from:
$29.99

$23.99

Lab 7: Normalization One Solution

Write some downright dif=icult SQL queries
Work really hard for lab points.
Check that your instance of our beloved CAP3 database is exactly the same as mine in the script on our class web site.
Use CAP3 to answer all of these questions.
1.
Display the name and city of customers who live in any city that makes the most
 
different kinds of products. (There are two cities that make the most different
 
products. Return the name and city of customers from either one of those.)
2.
Display the names of products whose priceUSD is strictly above the average priceUSD,
 
in reverse-alphabetical order.
3.
Display the customer name, pid ordered, and the total for all orders, sorted by total
 
from high to low.
4.
Display all customer names (in alphabetical order) and their total ordered, and
 
nothing more. Use coalesce to avoid showing NULLs.
5.
Display the names of all customers who bought products from agents based in Tokyo
 
along with the names of the products they ordered, and the names of the agents who
 
sold it to them.
6.
Write a query to check the accuracy of the dollars column in the Orders table. This
 
means calculating Orders.totalUSD from data in other tables and comparing those
 
values to the values in Orders.totalUSD. Display all rows in Orders where
 
Orders.totalUSD is incorrect, if any.
7.
What’s the difference between a LEFT OUTER JOIN and a RIGHT OUTER JOIN? Give
 
example queries in SQL to demonstrate. (Feel free to use the CAP2 database t make
 
your points here.)
Test, test, and test again. Then test some more. When you think you've tested enough, go
back and keep testing. Then get someone else to test for you while you test theirs.
Push your work to your GitHub repository early and often. Be sure to write meaningful
commit messages.

Chapter 6 in our text, especially 6.3 and 6.4
•  SQL tag at Stack Over=low - http://stackover=low.com/questions/tagged/sql

More products