No need to replicate, you just need to find equivalent which means producing same result on same database instance.
25 Nov 2018 - 10:59am
It’s absolutely fine. In fact it’s 0th level to 11th, so count is 12.
25 Nov 2018 - 10:57am
Yes, indeed there is. I am teaching this for last 3 years, and still get confused a lot.
20 Nov 2018 - 11:32am
As SQL provides you with the separate notion of handling NULL values using "IS NULL", "IS NOT NULL" etc. you should be using that. However, it's not mandatory to go into this trouble while using RA or TRC. They are theoretical languages, and their implementation can be handled separately while doing so in SQL.
20 Nov 2018 - 11:30am
Yes, to implement this in SQL, you must use "cname <> 'SBI' or cname IS NULL" in the where clause. However, as RA is a theoretical concept, we do write "a = NULL" or "a <> NULL" in RA SELECT clauses, which also does not produce correct result if used in SQL where clause.
20 Nov 2018 - 11:26am
NULL is not equal to ‘SBI’, so the tuples having NULL in cname should be selected.
20 Nov 2018 - 11:01am
For single implication as we were taking forall the courses taken by Rosie, similarly while using double implication, we should consider forall the courses taken by the other student.
20 Nov 2018 - 10:52am
And v should also have forall clause I think. And in the RA, we should go with subtraction, rather than division.
20 Nov 2018 - 9:45am
Thanks Samagra. Yes, you are right! I completely ignored the exactly same part. Replacing the implication by double implication will serve the purpose.
20 Nov 2018 - 9:39am
We assume relations by default have distinct tuples. If R and S both have distinct tuple, their natural join will also produce distinct tuples in this case.
20 Nov 2018 - 8:11am

Pages