Hi,

Is it ok for me to post some revision work and ask for it to be corrected here?
I am revising for a repeat exam atm and I have nobody to correct my work..

Here goes:

Question:
Table for question
Using the following example business rules and the example table of Appendix A (at the back of this exam paper):
List four functional dependencies using the standard notation you have studied. (You must refer to one of the above rules or state an assumption to qualify each functional dependency you list)
• Rule 1: A supplier number uniquely identifies a real-world supplier.
• Rule 2: A part number uniquely identifies a part.
• Rule 3: The (supplier number, part number) combination can be used to uniquely identify how a supplier supplies a given part (i.e. quantity supplied and any other details about a shipment)
• Rule 4: The city in which a supplier is located determines the status value for that supplier.

Answer:
sNum -> sName (Rule 1)
pNum-> pName (Rule 2)
sNum,pNum -> qtyDelivered (Rule 3)
sCity -> sStatus (Rule 4)

Fingers crossed..

Thanks Guys..this really means a lot. I've to repeat this exam early August and I have nobody else to correct my work..

You haven't told us what the question is. The final dependency in your answer isn't implied by any of the rules.

Nor the second.

pootle flump ur codings are working excelent.

My bad guys, sorry.

I did a bad copy and paste job.

I have corrected my answer also, that was also a silly copy error.

Thanks guys for looking at this.. I have an exam soon and have nobody to check my work.

Looks fine.

pootle flump ur codings are working excelent.

Thanks pootle_flump..

Ok, well here's a question onNormalisation that I did.. I'm pretty sure i'm correct but need some reassurance.

I'm happy with it but I'd just like to know for sure if i'm doing it right.. The amount of marks going or each question seems kind of odd.. 2 marks for part (i) even though I did a lot of work to get there.. Decompose table to 1NF and then 2NF. That brings me to my main query, am I doing too much here? Is it jst a one or two line answer that is required?
Part (ii) and part (iii) seem to be the same questions asked in slightly different ways but they have most amount of marks..

What am I missing here?

---------------------
Here's my work: (PK is underlined)
---------------------
Table:here

Questions:
Using the single-table of Appendix A as a start point
(i) Decompose an example of a 2NF table.(2 marks)
(ii) State why the table you specified as your answer to part (i), is in 2NF and not in 3NF.(4 marks)
(iii) Decompose an example of a 3NF table.(2 marks)
(iv) State why the table you specified as your answer to part (iii), is in 3NF and not in 2NF.(4 marks)

Answers:
(i)
1NF:
SingleTableDB(sNum,sName,sStatus,sCity,pNum,pName, pColour,pWeight,pQtyDelivered)

Dependencies:

sNum->sName,sStatus,sCity (Partial Dep)
pNum->pName,pWeight,pColour (Partial Dep)
sNum,pNum->pQtyDelivered (Full Dep)
sCity->sStatus (Transitive Dep)

Dependency Diagram: here

2NF:
Suppliers(sNum,sStatus,sCity)
Parts(pNum,pName,pWeight,pColour)
Delivieries(sNum,pNum,pQtyDelivered)

** Dependency Diagram for each new table here **
** Suppliers shows transitive dependency between sStatus and sCity **

(ii)
Because a transitive dependcy exists (???) Enough for 4 marks?

(iii)

sCity -> sStatus
Suppliers (sNum, sName, sCity)
Parts(pNum, pName, pColour, pWeight)
Deliveries(pNum,pName,pQtyDelivered)
Status(sCity,sStatus)

(iv)

Because the transitive dependency between sCity and sStatus no longer exists. I have made a new table Status that stores the relevant sStatus for each sCity. (???) Enough for 4 marks?

What is non-loss decomposition?
Non-Loss Decomposition happens during normalisation. It means that during normalisation (decomposing tables) no data is sacrificed due to the restructuring of tables.


Do you guys mind looking over this stuff??

Looks fine and I agree the marking appears imbalanced.
You did lose one column when you normalised to 2NF though....

pootle flump ur codings are working excelent.

Oop, forgot to put in sName.. Silly me, I do usually count up the number of attributes after each step.

Thanks for checking it over. It seems i'm not doing as bad as I had thought!

I suppose for ii) you could give details about specifically what the transitive dependency is rather than just saying one exists. Without even looking at the model anyone could come up with that answer to the question.

That marking scheme is really odd - ii and iV are effectively the same question.

pootle flump ur codings are working excelent.

Taken into consideration..thanks.

Thanks for you help.