CS Homework 8 p. 1. CS Homework 8

Transcription

1 CS Homework 8 p. 1 Deadline: 11:59 pm on Friday, October 27, 2017 Purpose: CS Homework 8 To practice normalizing sets of relations into 1NF, 2NF, and 3NF, to practice writing more nested selects/subselects, and to get more experience with SQL select statement order by clauses, group by clauses, and having clauses. How to submit Each time you wish to submit, within the directory 325hw8 on nrs-projects.humboldt.edu (and at the nrsprojects UNIX prompt, NOT inside sqlplus!) type: ~st10/325submit...to submit your current files, using a homework number of 8. (Make sure that the files you intend to submit are listed as having been submitted!) Additional notes: DB Reading Packet 6 and SQL Reading Packets 4 and 5, on the course Canvas site, along with the posted in-class projections from the public course web site, are useful references for this homework. Now that we have covered the order by clause, you are expected to use it appropriately when an explicit row ordering is specified. Queries for problems asking for explicit row ordering will be incorrect if they do not include a reasonable order by clause. You are required to use the HSU Oracle student database for Problem 2. An example hw8-out.txt has been posted along with this homework handout, to help you see if you are on the right track with your queries for Problem 2. If your hw8-out.txt matches this posted one, that doesn't guarantee that you wrote appropriate queries, but it is an encouraging sign. (I added a few extra prompt commands near the beginning of this script to output some blank lines for slightly-prettier output.) You are expected to follow course style standards for SQL and SQL*Plus statements. AND, we are ADDING the following ADDITIONAL course style standards for SQL select statements: It is considered poor style to deliberately include a Cartesian product without adding the appropriate number of join conditions to make it an equi-join. (That is, if you have n tables in a from clause, you are expected to also have (n-1) join conditions.) An order by clause should not be used in a sub-select; it should only be used in an outermost select.

2 CS Homework 8 p. 2 When a select statement includes an order by clause, order by should be the final clause, lined up with the rest of the outermost select. You should only use group by for a reason (usually, because you want some computation for the rows in each group). If you aren't performing some computation on the rows in each group, do not use group by. In particular, don't use it just to suppress duplicate rows -- that is what DISTINCT is for! Problem 1 Use ssh to connect to nrs-projects.humboldt.edu, and create, protect, and go to a directory named 325hw8 on nrs-projects: mkdir 325hw8 chmod hw8 cd 325hw8 Put all of your files for this homework in this directory. (And it is from this directory that you should type ~st10/325submit to submit your files each time you want to submit the work you have done so far.) Use nano (or vi or emacs) to create a file named hw8-1.txt: nano hw8-1.txt While within nano (or vi or emacs), type in your name, and then an answer for each of the following, preceding each answer with the number of the question being answered. Normalization Scenario 1 This is a scenario in which volunteers work on different projects, and those within the scenario keep track of the work volunteers do for different projects. Here is a relation and additional functional dependency information: Volunteer_Record(VOL_NUM, PROJ_NUM, vol_lname, vol_fname, proj_name, total_hrs_worked, task_type_code, task_type_descr) vol_num -> vol_lname, vol_fname proj_num -> proj_name task_type_code -> task_type_descr Problem 1-1 Is this relation already in 1NF? If not, do what is necessary so that it is; if so, leave it as-is. Write the relation structure(s) for the 1NF relation(s) in either case. (ONLY make those changes, if needed, necessary to reach 1NF -- do NOT go beyond, yet!) Remember to indicate foreign keys, if any, using SQL foreign key syntax directly under the relation structure that has that foreign key.

3 CS Homework 8 p. 3 Problem 1-2 Are your relation(s) resulting from Problem 1-1 in 2NF? If not, do what is necessary so that is is/they are; if Write the relation structure(s) for the 2NF relation(s) in either case (including those that are unchanged, if any.) (ONLY make those changes, if needed, necessary to reach 2NF -- do NOT go beyond, yet!) Problem 1-3 Are your relation(s) resulting from Problem 1-2 in 3NF? If not, do what is necessary so that is is/they are; if Write the relation structure(s) for the 3NF relation(s) in either case (including those that are unchanged, if any.) (ONLY make those changes, if needed, necessary to reach 3NF -- do NOT go beyond, yet!) Normalization Scenario 2 Now consider this second scenario -- note that this time, this is deliberately generic, so that I can see if you are getting the normalization steps without any extraneous semantic information accidentally getting in the way. All of the information that you need is below if you understand the normalization process as described in lecture and in DB Reading Packet 6. Here is a relation and additional functional dependency information: R1 (A, B, C, D, e, f, g, h, i, j, k, l, m, n, o) (e, f) -> (k, l) (a, c) -> o h -> m d -> n NOTE --- "Whoops!" An "error" was made above --- even though, as written above, we know that it should be true that (a, b, c, d) --> g, it is actually the case that attribute g can have multiple values for a particular set of values (a, b, c, d). (That is, g is a multi-valued attribute.) (Note that it needs all 4 to determine those multiple values of g, however.) Rhetorical question (you do not have to answer in your homework file): what does that then imply about R1 above? Problem 1-4 Is this relation already in 1NF? If not, do what is necessary so that it is; if so, leave it as-is. Write the relation structure(s) for the 1NF relation(s) in either case; you do not need to re-write the functional dependencies. (ONLY make those changes, if needed, necessary to reach 1NF -- do NOT go beyond, yet!) Remember to indicate foreign keys, if any, using SQL foreign key syntax directly under the relation structure that has that foreign key.

4 CS Homework 8 p. 4 Problem 1-5 Are your relation(s) resulting from Problem 1-4 in 2NF? If not, do what is necessary so that is is/they are; if Write the relation structure(s) for the 2NF relation(s) in either case (including those that are unchanged, if any); you do not need to re-write the functional dependencies. (ONLY make those changes, if needed, necessary to reach 2NF -- do NOT go beyond, yet!) Problem 1-6 Are your relation(s) resulting from Problem 1-5 in 3NF? If not, do what is necessary so that is is/they are; if Write the relation structure(s) for the 3NF relation(s) in either case (including those that are unchanged, if any); you do not need to re-write the functional dependencies. (ONLY make those changes, if needed, necessary to reach 3NF -- do NOT go beyond, yet!) Problem 2 This problem again uses the tables created by the SQL script movies-create.sql and populated by movies-pop.sql. As a reminder, these tables can be described in relation structure form as: Movie_category(CATEGORY_CODE, category_name) Client(CLIENT_NUM, client_lname, client_fname, client_phone, client_credit_rtg, client_fave_cat) foreign key (client_fave_cat) references movie_category(category_code) Movie(MOVIE_NUM, movie_title, movie_director_lname, movie_yr_released, movie_rating, category_code) foreign key(category_code) references movie_category Video(VID_ID, vid_format, vid_purchase_date, vid_rental_price, movie_num) foreign key (movie_num) references movie Rental(RENTAL_NUM, client_num, vid_id, date_out, date_due, date_returned) foreign key (client_num) references client, foreign key(vid_id) references video And, again, for your convenience as a reference, a handout of these relation structures is posted along with this homework handout. (These tables should still exist in your database from Homework 5, so you should not need to re-run movies-create.sql or movies-pop.sql unless you have been experimenting with insertions or other table modifications.) Use nano (or vi or emacs) to create a file named hw8.sql: nano hw8.sql While within nano (or whatever), type in the following within one or more SQL comments: your name

5 CS Homework 8 p. 5 CS Homework 8 - Problem 2 the date this file was last modified Then: use spool to start writing the results for this script's actions into a file hw8-out.txt put in a prompt command printing Homework 8 Problem 2 put in a prompt command printing your name include a spool off command, at the BOTTOM/END of this file. Type your answers to the problems below BEFORE this spool off command! NOTE!!! READ THIS!!! Now, within your file hw8.sql, add in SQL statements for the following, PRECEDING EACH with a SQL*Plus prompt command noting what problem part it is for. Problem 2-1 Perform a relational selection of the rows of the client table, displaying the resulting rows in increasing order of client credit rating. Then, perform another relational selection of the rows of the client table, but now displaying the resulting rows in decreasing order of client credit rating. Problem 2-2 On a previous homework, you wrote a query which projects one column in its result: this column, with heading "Rating: Movie", shows, for each movie, the rating for that movie, then a colon and a space, and then the title of that movie. Now write a version of this query so that, now, its results are ordered by increasing/alphabetical order of movie rating, and for rows with the same movie rating, they should be ordered by increasing/alphabetical order of movie title. Problem 2-3 Perform a projection of the name of a movie's category, the movie title, and the movie rating, for all movies, displaying the resulting rows in order of movie rating, and for movies with the same rating, in reverse alphabetical order of move category name, and for movies with the same rating and category name, in order of movie title. Problem 2-4 Project the client's last name, telephone number, and credit rating for clients whose credit rating is equal to or higher than the average client credit rating, displaying the resulting rows in reverse order of credit rating. Problem 2-5 From the video table, for each video format, project the video format, the number of videos with that format using the column alias QTY, and the average video rental price for videos with that format using the column

6 CS Homework 8 p. 6 alias AVG RENTAL PRICE. (Do not worry about the ugly formatting of the average video rental price.) Problem 2-6 Rewrite Problem 2-5's query, this time displaying the resulting rows in increasing order of the number of videos with that format. Problem 2-7 From the video table, for each video rental price, project the video rental price, and the number of videos with that rental price using the column alias # VIDS, displaying the resulting rows in decreasing order of video rental price. Problem 2-8 Rewrite Problem 2-7's query, except this time include only those video rental prices that are the prices of at least 5 (5 or more) videos. (That is, project the video rental price and number of videos with that rental price using the column alias # VIDS ONLY for video rental prices that are the prices of 5 or more videos.) Problem 2-9 From the movie and movie_category tables, for each movie category, project the movie category name, and the number of movies in that category using the column alias # MOVIES. Display the resulting rows in reverse order of the number of movies in each category(that is, display the row with the category with the most movies first). Problem 2-10 Project the average number of videos per video rental price, using the column alias AVG # PER PRICE. (This is a single value, note.) (Hint: use an aggregate function with an aggregate function as its argument...) Problem 2-11 Remember that, on Homework 7, Problem 2-2, you projected the video id's only of all videos that have a rental price less than the average video rental price. Let's build atop that query. Using just sub-selects, and using NO join or Cartesian product operations, project the client numbers only for clients involved in rentals of videos that have a rental price less than the average video rental price. Do not worry, in this problem, about whether or not there are duplicate rows in the result. Problem 2-12 And, taking this one step further... Using just sub-selects, and using NO join or Cartesian product operations, project the last names and the client credit ratings of clients involved in rentals of videos that have a rental price less than the average video rental price. (If done correctly, there is no way that you can get duplicate rows in this query's results, even without the keyword that prevents them -- do you understand why? You do not have to answer as part of this homework, but you should know...)

7 CS Homework 8 p. 7 Problem 2-13 Remember that, on Homework 7, Problem 2-3, you were asked to use an all-nesting approach to project the last names and first names only for clients who have ever rented the video with ID '130012'. Now I want you to write a slightly-expanded version of this query using a DIFFERENT approach: Now using a join, and NOT using ANY nesting or sub-selects, project the last names, first names, and date the video was due for clients who have ever rented the video with ID '130012'. Problem 2-14 Now, I don't want you to get the mistaken impression that we must do EITHER joins OR sub-selects -- a query can also involve both! Project the last names, favorite movie category names, and credit ratings for clients who have credit ratings higher than the average credit rating for all clients. (NOTE: I am not asking you to project the client_fave_cat --- I am asking you to project the name of the category corresponding to the client_fave_cat.) HINT: a select clause can only project attributes from relation(s) in its corresponding from clause Problem 2-15 Using NOT EXISTS, write a query that will project the number of videos which have never been rented and their average rental price. (This will not be accepted as correct unless it properly uses NOT EXISTS.) Problem 2-16 Using NOT EXISTS, for videos that have never been rented, project the video id's, the title of the movie on that video, and the format of that video. (This will not be accepted as correct unless it properly uses NOT EXISTS.) When you think the results of all of these queries look correct, this would also be a good time to look at the contents of hw8-out.txt -- at the nrs-projects prompt (the UNIX level, NOT in sqlplus!), type: more hw8-out.txt You should see that hw8-out.txt contains the query results you just saw within sqlplus. (And, remember that you can compare your hw8-out.txt contents to those in the posted example hw8- out.txt posted along with this homework handout, to help you see if you are on the right track with your answers, although remember to also meet each problem's stated requirements and the class style standards.) When you are satisfied with these, then hw8.sql and hw8-out.txt are completed.