How Can Relational Algebra Be Used to Find Diseases With Only One Medication?

[Shaitan00]

I am trying to write the RELATIONAL ALGEBRA statement for a query defined as follows:
"Find the diseases for which there is only one medication"

I have a table TREATMENT with two attributes, DISEASE & MEDICATION
TREATMENT (disease, medication)

This table holds a list of diseases and corresponding medication, of course for each diseases that can potentially be multiple medications, and for some only one. For example assume you have something that resembles the following:

Disease | Medication
--------------------
FLU | med1
FLU | med2
CANCER | med3
HIV | med4
FLU | med5

In this case you would expect to get CANCER & HIV (as they only have 1 corresponding medication) whereas FLU would be excluded because it has 3... (etc...)

Any help would be greatly appreciated...
Thanks,

 

[dodo]

See "aggregate functions", around halfway this page:
http://www.databasteknik.se/webbkursen/relalg-lecture/index.html

And this is probably better in the computer science forums...

 

[Shaitan00]

I was looking at that already - but couldn't figure it out...
WIth those expressions I can easily determine the COUNT for each but how do I list only the diseases that have a count = 1?

Thanks,

 

[dodo]

SQL uses a HAVING keyword for such cases, which is nothing but a select operation on the result of the aggregation. I believe the only reason to define in SQL a different keyword than WHERE was to distinguish between a select operation after the aggregation (HAVING), as opposed to before (WHERE).