So I want to explain a couple of other key database terms. One of those is cardinality, cardinality has two meanings. One as it refers to table relationships, cardinality refers to a one to one, one to many, many to one type relationship. Cardinality refers to the relationships between and among tables. It can also mean unique values in a column. A column that has a low cardinality really means that it's a column that has a few unique values. An example of a column that has few unique values, Would be things such as gender, would be things such as status paid, unpaid, those are a few unique values. A table can consist of a column called status and that status column will have four or five perhaps values, possible values across the entire table. A column with high cardinality refers to a column that has a lot of unique values. Primary keys are very high cardinality. User name, very high cardinality because you're not going to have user names that are repeated across multiple rows in a given column. High cardinality, very unique. Low cardinality not unique at all. Then there's something called Cartesian joint. Let me show you this, I have a table called DEPTS, short for Departments and a table here called Employees. Notice I have a department ID in the employees table and I have an emp ID in the employees table. I have a DEPT ID or department ID in the DEPTS table. I go over here and I do a select everything from the, DEPT table, and I take a look at what that produces. If you look here, I've got three rows, Department ID, Dept ID 123, Sales, Marketing and Tech Support. If I do the same thing and take a look at the employees table. Like so. And I run this, I've got three rows in the employee table, employees table has three rows and my DEPT table has three rows. If I do a Cartesian join, if I just do a ,DEPTS, what's going to happen is just going to produce nine rows, nine rows. Look what happens, I joined the table without really specifying what I'm joining it with, Cartesian join is usually a problem with your query, look what it did. It put John Adams in Sales and in Marketing and in Tech Support. And then it put James Monroe in Sales, and Marketing, and Tech Support. And it gave me Serena Williams in Sales, Marketing and Tech Support. Basically the three names and the three departments, it basically does a multiplier, three times three giving me nine possible results, which is not what I want. What I really wanted to do was I wanted to join the employees table with the department table based on the department, see this department ID. Okay, so if I don't join them, I get a Cartesian join and if I do a depth like that be and then I say where A.DEPTID equals B.DEPTID. Now, I'm joining them together. And now I get John is in sales, James is in Marketing, and Serena is in Tech Support. Okay, so this is not a Cartesian join because I'm actually using a proper join, I'm joining the DEPT ID on both sides, okay? Whereas if I don't do a proper joint and I just combine the two, this is combining the two. Instead of getting three results, I get nine, three times three. The number of total rows in the employees table times the number of total rows in the DEPT table is what gives me the Cartesian join. Cartesian join are basically you're combining the two tables and without having to properly join them together, Cartesian join and cardinality.
