Add postgres day 2 homework

peferron · Nov 6, 2016 · 3e8241e · 3e8241e
1 parent ad56940
commit 3e8241e
Show file tree

Hide file tree

Showing 3 changed files with 329 additions and 13 deletions.
diff --git a/README.md b/README.md
@@ -1,2 +1,3 @@
 # 7dbs-in-7wks
-Code from the "Seven Databases in Seven Weeks" book.
+
+Source code from the book is available on the [official website](https://pragprog.com/book/rwdata/seven-databases-in-seven-weeks).
diff --git a/postgres/.gitignore b/postgres/.gitignore
@@ -0,0 +1 @@
+*.sql
diff --git a/postgres/homework.md b/postgres/homework.md
@@ -2,26 +2,340 @@
 
 ## Day 1
 
-1. Select all the tables we created (and only those) from `pg_class`.
+
+### 1.
+
+```
+SELECT relname FROM pg_class
+WHERE relname !~ '^(pg_|sql_)'
+AND relkind = 'r';
+```
+
+### 2.
+
+```
+SELECT c.country_name FROM countries c
+JOIN venues v ON c.country_code = v.country_code
+JOIN events e ON v.venue_id = e.venue_id
+WHERE e.title = 'LARP Club';
+```
+
+### 3.
+
+```
+ALTER TABLE venues
+ADD COLUMN active BOOLEAN DEFAULT TRUE;
+```
+
+## Day 2
+
+### 1.
+
+```
+CREATE RULE delete_venue
+AS ON DELETE TO venues DO INSTEAD
+UPDATE venues SET active = false WHERE venue_id = OLD.venue_id;
+```
+
+### 2.
+
+```
+SELECT * FROM generate_series(1, 12);
+```
+
+### 3.
+
+This exercise is fairly difficult. Here are the steps I followed.
+
+1. We want a final result like this (`wom` is the index of the week in the month):
 
  ```
- SELECT relname FROM pg_class
- WHERE relname !~ '^(pg_|sql_)'
- AND relkind = 'r';
+ wom | sun | mon | tue | wed | thu | fri | sat
+ -----+-----+-----+-----+-----+-----+-----+-----
+ 0 | | | | | | |
+ 1 | | | | | | |
+ 2 | | | 1 | 1 | | |
+ 3 | 2 | | | | | |
+ 4 | | | | | | |
  ```
+
 
-2. Write a query that finds the country name of the LARP Club event.
+2. We know a `crosstab` can achieve the above if we feed it a table like this:
 
  ```
- SELECT c.country_name FROM countries c
- JOIN venues v ON c.country_code = v.country_code
- JOIN events e ON v.venue_id = e.venue_id
- WHERE e.title = 'LARP Club';
+ wom | dow | count
+ -----+-----+-------
+ 0 | 3 |
+ 0 | 4 |
+ 0 | 5 |
+ 0 | 6 |
+ 0 | 0 |
+ 1 | 1 |
+ 1 | 2 |
+ 1 | 3 |
+ 1 | 4 |
+ 1 | 5 |
+ 1 | 6 |
+ 1 | 0 |
+ 2 | 1 |
+ 2 | 2 | 1
+ 2 | 3 | 1
+ 2 | 4 |
+ 2 | 5 |
+ 2 | 6 |
+ 2 | 0 |
+ 3 | 1 |
+ 3 | 2 |
+ 3 | 3 |
+ 3 | 4 |
+ 3 | 5 |
+ 3 | 6 |
+ 3 | 0 | 2
+ 4 | 1 |
+ 4 | 2 |
+ 4 | 3 |
  ```
+
+3. So we need to start by listing all the days in the month (we'll pick February 2012 since it has the most events):
 
-3. Alter the `venues` table to contain a boolean column called `active`, with the default value of `true`.
+ ```
+ SELECT date
+ FROM generate_series(
+ '2012-02-01'::date,
+ '2012-02-29'::date,
+ '1 day'::interval
+ ) AS date;
+ ```
+  
+ ```
+ date
+ ------------------------
+ 2012-02-01 00:00:00+01
+ 2012-02-02 00:00:00+01
+ 2012-02-03 00:00:00+01
+ 2012-02-04 00:00:00+01
+ 2012-02-05 00:00:00+01
+ 2012-02-06 00:00:00+01
+ 2012-02-07 00:00:00+01
+ 2012-02-08 00:00:00+01
+ 2012-02-09 00:00:00+01
+ 2012-02-10 00:00:00+01
+ 2012-02-11 00:00:00+01
+ 2012-02-12 00:00:00+01
+ 2012-02-13 00:00:00+01
+ 2012-02-14 00:00:00+01
+ 2012-02-15 00:00:00+01
+ 2012-02-16 00:00:00+01
+ 2012-02-17 00:00:00+01
+ 2012-02-18 00:00:00+01
+ 2012-02-19 00:00:00+01
+ 2012-02-20 00:00:00+01
+ 2012-02-21 00:00:00+01
+ 2012-02-22 00:00:00+01
+ 2012-02-23 00:00:00+01
+ 2012-02-24 00:00:00+01
+ 2012-02-25 00:00:00+01
+ 2012-02-26 00:00:00+01
+ 2012-02-27 00:00:00+01
+ 2012-02-28 00:00:00+01
+ 2012-02-29 00:00:00+01
+ ```
+
+2. Then join with the events table:
 
  ```
- ALTER TABLE venues
- ADD COLUMN active BOOLEAN DEFAULT TRUE;
+ SELECT date, title
+ FROM (
+ SELECT date
+ FROM generate_series(
+ '2012-02-01'::date,
+ '2012-02-29'::date, 
+ '1 day'::interval
+ ) AS date
+ ) AS dates
+ LEFT JOIN events
+ ON date = date_trunc('day', starts);
+ ```
+  
+ ```
+ date | title
+ ------------------------+-----------------
+ 2012-02-01 00:00:00+01 |
+ 2012-02-02 00:00:00+01 |
+ 2012-02-03 00:00:00+01 |
+ 2012-02-04 00:00:00+01 |
+ 2012-02-05 00:00:00+01 |
+ 2012-02-06 00:00:00+01 |
+ 2012-02-07 00:00:00+01 |
+ 2012-02-08 00:00:00+01 |
+ 2012-02-09 00:00:00+01 |
+ 2012-02-10 00:00:00+01 |
+ 2012-02-11 00:00:00+01 |
+ 2012-02-12 00:00:00+01 |
+ 2012-02-13 00:00:00+01 |
+ 2012-02-14 00:00:00+01 | Valentine's Day
+ 2012-02-15 00:00:00+01 | LARP Club
+ 2012-02-16 00:00:00+01 |
+ 2012-02-17 00:00:00+01 |
+ 2012-02-18 00:00:00+01 |
+ 2012-02-19 00:00:00+01 |
+ 2012-02-20 00:00:00+01 |
+ 2012-02-21 00:00:00+01 |
+ 2012-02-22 00:00:00+01 |
+ 2012-02-23 00:00:00+01 |
+ 2012-02-24 00:00:00+01 |
+ 2012-02-25 00:00:00+01 |
+ 2012-02-26 00:00:00+01 | Wedding
+ 2012-02-26 00:00:00+01 | Dinner with Mom
+ 2012-02-27 00:00:00+01 |
+ 2012-02-28 00:00:00+01 |
+ 2012-02-29 00:00:00+01 |
+ ```
+
+3. Show the count instead of the title:
+
+ ```
+ SELECT date, NullIf(count(events), 0) AS count
+ FROM (
+ SELECT date
+ FROM generate_series(
+ '2012-02-01'::date,
+ '2012-02-29'::date, 
+ '1 day'::interval
+ ) AS date
+ ) AS dates
+ LEFT JOIN events
+ ON date = date_trunc('day', starts)
+ GROUP BY date
+ ORDER BY date;
+ ```
+  
+ ```
+ date | count
+ ------------------------+-------
+ 2012-02-01 00:00:00+01 |
+ 2012-02-02 00:00:00+01 |
+ 2012-02-03 00:00:00+01 |
+ 2012-02-04 00:00:00+01 |
+ 2012-02-05 00:00:00+01 |
+ 2012-02-06 00:00:00+01 |
+ 2012-02-07 00:00:00+01 |
+ 2012-02-08 00:00:00+01 |
+ 2012-02-09 00:00:00+01 |
+ 2012-02-10 00:00:00+01 |
+ 2012-02-11 00:00:00+01 |
+ 2012-02-12 00:00:00+01 |
+ 2012-02-13 00:00:00+01 |
+ 2012-02-14 00:00:00+01 | 1
+ 2012-02-15 00:00:00+01 | 1
+ 2012-02-16 00:00:00+01 |
+ 2012-02-17 00:00:00+01 |
+ 2012-02-18 00:00:00+01 |
+ 2012-02-19 00:00:00+01 |
+ 2012-02-20 00:00:00+01 |
+ 2012-02-21 00:00:00+01 |
+ 2012-02-22 00:00:00+01 |
+ 2012-02-23 00:00:00+01 |
+ 2012-02-24 00:00:00+01 |
+ 2012-02-25 00:00:00+01 |
+ 2012-02-26 00:00:00+01 | 2
+ 2012-02-27 00:00:00+01 |
+ 2012-02-28 00:00:00+01 |
+ 2012-02-29 00:00:00+01 |
+ ```
+
+4. Show the week of month and day of week instead of the date:
+
+ ```
+ SELECT
+ extract(week from date) - extract(week from date_trunc('month', date)) AS wom,
+ extract(dow from date) AS dow,
+ NullIf(count(events), 0) AS count
+ FROM (
+ SELECT date
+ FROM generate_series(
+ '2012-02-01'::date,
+ '2012-02-29'::date, 
+ '1 day'::interval
+ ) AS date
+ ) AS dates
+ LEFT JOIN events
+ ON date = date_trunc('day', starts)
+ GROUP BY date
+ ORDER BY date;
+ ```
+  
+ ```
+ wom | dow | count
+ -----+-----+-------
+ 0 | 3 |
+ 0 | 4 |
+ 0 | 5 |
+ 0 | 6 |
+ 0 | 0 |
+ 1 | 1 |
+ 1 | 2 |
+ 1 | 3 |
+ 1 | 4 |
+ 1 | 5 |
+ 1 | 6 |
+ 1 | 0 |
+ 2 | 1 |
+ 2 | 2 | 1
+ 2 | 3 | 1
+ 2 | 4 |
+ 2 | 5 |
+ 2 | 6 |
+ 2 | 0 |
+ 3 | 1 |
+ 3 | 2 |
+ 3 | 3 |
+ 3 | 4 |
+ 3 | 5 |
+ 3 | 6 |
+ 3 | 0 | 2
+ 4 | 1 |
+ 4 | 2 |
+ 4 | 3 |
+ ```
+
+5. We're now ready for the `crosstab`:
+
+ ```
+ SELECT * from crosstab(
+ 'SELECT
+ extract(week from date) - extract(week from date_trunc(''month'', date)) AS wom,
+ extract(dow from date) AS dow,
+ NullIf(count(events), 0) AS count
+ FROM (
+ SELECT date
+ FROM generate_series(
+ ''2012-02-01''::date,
+ ''2012-02-29''::date, 
+ ''1 day''::interval
+ ) AS date
+ ) AS dates
+ LEFT JOIN events
+ ON date = date_trunc(''day'', starts)
+ GROUP BY date
+ ORDER BY date', 
+ 'SELECT * from generate_series(0, 6)'
+ ) AS (
+ wom int,
+ sun int, mon int, tue int, wed int, thu int, fri int, sat int
+ )
+ ORDER BY wom;
+ ```
+
+ ```
+ wom | sun | mon | tue | wed | thu | fri | sat
+ -----+-----+-----+-----+-----+-----+-----+-----
+ 0 | | | | | | |
+ 1 | | | | | | |
+ 2 | | | 1 | 1 | | |
+ 3 | 2 | | | | | |
+ 4 | | | | | | |
+ (5 rows)
  ```
+
+ Yay!