+ 2

Can anyone explain why Q1, Q2, Q3 will gives the same reason

In the following, the schema of R is R(a,b). Q1: SELECT * FROM R Q2: (SELECT * FROM R) INTERSECT (SELECT* FROM R); Q3: SELECT DISTINCT * FROM R; They can have duplicate values too can anyone explain

2nd Oct 2019, 9:47 AM
Preity
Preity - avatar
8 odpowiedzi
+ 2
If R contains duplicates then Q1 and Q3 are clearly different, compare: SELECT * FROM ( SELECT 1,2 UNION ALL 2,3 UNION ALL 2,3 ) R Result: 1,2 2,3 2,3 with SELECT DISTINCT * FROM ( SELECT 1,2 UNION ALL 2,3 UNION ALL 2,3 ) R Result: 1,2 2,3 However, if the key is `ab`, then all rows are unique by definition and there are no duplicates. If the key is `a` then there can be no two columns with the same `a` and `b` values either so there are no duplicates. Same goes for when the key is `b`.
2nd Oct 2019, 10:46 AM
Schindlabua
Schindlabua - avatar
+ 1
Schindlabua 12 23 23 Not possible as same row value not possible in rdbms, or did I miss your logic
2nd Oct 2019, 11:39 AM
Preity
Preity - avatar
0
Q1 is take everything from R Q2 is using intersect, it'll return the data if. both left and right query have/share that same data. since both query are same, the result will be same thus everything will be return. (i'm not sure what happem if there's a duplicate rows inside the table(s) ) Q3 is same as Q1 but the distinct operation will remove any duplicate data from the result instead of showing everything
2nd Oct 2019, 10:44 AM
Taste
Taste - avatar
0
Preity Not in theory because relations are sets and set members are unique. In practice you will usually have a key which will prevent duplicates from happening. However given no key both MYSQL nor MSSQL happily accept duplicate rows: CREATE TABLE R(a integer, b integer); INSERT INTO R(1,2); INSERT INTO R(1,2); SELECT * FROM R; In your question you said "They can have duplicate values too" so I thought that is what you meant.
2nd Oct 2019, 11:44 AM
Schindlabua
Schindlabua - avatar
0
Schindlabua I'm solving this question so in that information is missing. But yes I want to know if duplicate exist then what the answer effect on the query. in this question given answer is option a and I'm getting C that's confusing https://www.sololearn.com/post/155300/?ref=app
2nd Oct 2019, 11:46 AM
Preity
Preity - avatar
0
If you applied distinct in all attributes(means distinct *) of table . There is no meaning of that because in rdbms table all rows ( tuples) are unique. So the third query will affect much or not
2nd Oct 2019, 11:49 AM
Preity
Preity - avatar
0
Preity Yeah you got it. This is where theory differs from practice. All attributes of a relation always form a superkey which also means there can be no two rows that are the same. In this case DISTINCT does nothing because each row is unique anyway. Seeing how this is a textbook question you would be right in saying Q1, Q2, Q3 do the same.
2nd Oct 2019, 11:54 AM
Schindlabua
Schindlabua - avatar
0
Schindlabua cool thanks for the help. Taste thanks for your comment too
2nd Oct 2019, 12:07 PM
Preity
Preity - avatar