Dr. Sun, the Extendable Hash Index example in your slides uses binary keys (e.g. 1011). Meanwhile, the example in the sample final uses decimal keys (e.g. 15). So if we are given decimal keys on the final, then would we need to convert them into binary first, before we insert them into the index? I ask also because you talk about "rightmost bits" in your slides, rather than "rightmost digits".

Also, it seems like we don't use the hash function at all when we insert keys into the Extendable Hash Index. For example, in your slides, you gave us the hash function K % 2^4, but we don't use it. We use the bucket directory instead. So how does the hash function fit in here?

I don't know if were allowed to post links on here, but this youtube video explained it clearly for me: LNK

Depending on you Global Depth, you use mod 2^g where g is the global depth.  So lets say the global depth is 2, you get the number using 2^2 = 4.  So, lets say you have your Bucket Directory:

Bucket Dir                                   Bucket Block

00

01

11

10                                                        14

Now when you do mod 4 of the decimal you are trying to add you match up the result to the binary number.  EX. 14 mod 4 = 2.  So 14 would go toe the Bucked assigned for binary number 2 (10 = 2).  This is just an example, in reality, you should already have more numbers in your buckets but this is just to show you how to insert a decimal number.

Now if your global depth is 3, you do 2^3 = 8.   So, now you do your decimal number mod 8.  and match the result to your bucket directory.

Sun, please correct me if I am wrong here.