The document discusses a SQL query that calculates the average product quality for sequential groups of the same product in a table. The table contains product test data with a batch number, product code, and quality score. The query uses self joins to group batches of the same product together, and then calculates the average quality for each group. It provides the sample data, the full SQL query, and step-by-step explanations of how the query works. The goal is to group contiguous batches of the same product together, while treating batches that are not contiguous as separate groups.

1. プログラマのためのSQL
P492の難解SQL

2. テーブル定義
CREATE TABLE ProductTests
(batch_nbr INTEGER NOT NULL PRIMARY KEY,
prod_code CHAR(1) NOT NULL,
prod_quality DECIMAL(8,4) NOT NULL);
INSERT INTO ProductTests (batch_nbr, prod_code, prod_quality)
VALUES (1, 'A', 80),
(2, 'A', 70),
(3, 'A', 80),
(4, 'B', 60),
(5, 'B', 90),
(6, 'C', 80),
(7, 'D', 80),
(8, 'A', 50),
(9, 'C', 70);

3. 問題のクエリ
SELECT X.prod_code, MIN(X.batch_nbr) AS start_batch_nbr, end_batch_nbr,
AVG(B4.prod_quality) AS avg_prod_quality
FROM (SELECT B1.prod_code, B1.batch_nbr,
MAX(B2.batch_nbr) AS end_batch_nbr
FROM ProductTests AS B1, ProductTests AS B2
WHERE B1.batch_nbr <= B2.batch_nbr
AND B1.prod_code = B2.prod_code
AND B1.prod_code
= ALL (SELECT prod_code
FROM ProductTests AS B3
WHERE B3.batch_nbr BETWEEN B1.batch_nbr
AND B2.batch_nbr)
GROUP BY B1.prod_code, B1.batch_nbr) AS X
INNER JOIN
ProductTests AS B4
ON B4.prod_code = X.prod_code
AND B4.batch_nbr = X.batch_nbr
GROUP BY X.prod_code, X.end_batch_nbr;

4. 何がしたいクエリなのか？
製品のリストと品質スコアのデータを持つProductTestsというテーブルを考
えよう。このテーブルから、同じ製品のシーケンシャルなグループに対する品
質スコアの平均を求めたい。たとえば、製品Aのバッチ番号1、2、3に対する
平均品質スコアは、
(80+70+80)/3=76.6666…
といった具合だ。ここで注意が必要なのは、バッチ番号8番の製品Aのように、
飛び石になった場合は別グループとして扱うことだ。
Joe Celko. プログラマのためのSQL 第4版 すべてを知り尽くしたいあなたに

5. 対象データと結果
対象データ
select * from ProductTests;
batch_nbr | prod_code | prod_quality
-----------+-----------+--------------
1 | A | 80.0000
2 | A | 70.0000
3 | A | 80.0000
4 | B | 60.0000
5 | B | 90.0000
6 | C | 80.0000
7 | D | 80.0000
8 | A | 50.0000
9 | C | 70.0000
(9 rows)
件のSQLの実行結果
prod_code | start_batch_nbr | end_batch_nbr | avg_prod_quality
-----------+-----------------+---------------+---------------------
A | 1 | 3 | 76.6666666666666667
B | 4 | 5 | 75.0000000000000000
C | 6 | 6 | 80.0000000000000000
D | 7 | 7 | 80.0000000000000000
A | 8 | 8 | 50.0000000000000000
C | 9 | 9 | 70.0000000000000000
(6 rows)

6. クエリ解説1※実行SQLは赤字で示します
SELECT X.prod_code, MIN(X.batch_nbr) AS start_batch_nbr, end_batch_nbr,
AVG(B4.prod_quality) AS avg_prod_quality
FROM (SELECT B1.prod_code, B1.batch_nbr,
MAX(B2.batch_nbr) AS end_batch_nbr
FROM ProductTests AS B1, ProductTests AS B2
WHERE B1.batch_nbr <= B2.batch_nbr
AND B1.prod_code = B2.prod_code
AND B1.prod_code
= ALL (SELECT prod_code
FROM ProductTests AS B3
WHERE B3.batch_nbr BETWEEN B1.batch_nbr
AND B2.batch_nbr)
GROUP BY B1.prod_code, B1.batch_nbr) AS X
INNER JOIN
ProductTests AS B4
ON B4.prod_code = X.prod_code
AND B4.batch_nbr = X.batch_nbr
GROUP BY X.prod_code, X.end_batch_nbr;
※注意1 SELECT句の列指定は*としました。
※注意2 わかりやすさのためOrder byを加えています(行順意外に結果はありません)
batch_nbr | prod_code | prod_quality | batch_nbr | prod_code | prod_quality
-----------+-----------+--------------+-----------+-----------+--------------
1 | A | 80.0000 | 1 | A | 80.0000
1 | A | 80.0000 | 2 | A | 70.0000
1 | A | 80.0000 | 3 | A | 80.0000
1 | A | 80.0000 | 8 | A | 50.0000
2 | A | 70.0000 | 2 | A | 70.0000
2 | A | 70.0000 | 3 | A | 80.0000
2 | A | 70.0000 | 8 | A | 50.0000
3 | A | 80.0000 | 3 | A | 80.0000
3 | A | 80.0000 | 8 | A | 50.0000
4 | B | 60.0000 | 4 | B | 60.0000
4 | B | 60.0000 | 5 | B | 90.0000
5 | B | 90.0000 | 5 | B | 90.0000
6 | C | 80.0000 | 6 | C | 80.0000
6 | C | 80.0000 | 9 | C | 70.0000
7 | D | 80.0000 | 7 | D | 80.0000
8 | A | 50.0000 | 8 | A | 50.0000
9 | C | 70.0000 | 9 | C | 70.0000
(17 rows)

7. ※実行SQLは赤字で示します
SELECT X.prod_code, MIN(X.batch_nbr) AS start_batch_nbr, end_batch_nbr,
AVG(B4.prod_quality) AS avg_prod_quality
FROM (SELECT B1.prod_code, B1.batch_nbr,
MAX(B2.batch_nbr) AS end_batch_nbr
FROM ProductTests AS B1, ProductTests AS B2
WHERE B1.batch_nbr <= B2.batch_nbr
AND B1.prod_code = B2.prod_code
AND B1.prod_code
= ALL (SELECT prod_code
FROM ProductTests AS B3
WHERE B3.batch_nbr BETWEEN B1.batch_nbr
AND B2.batch_nbr)
GROUP BY B1.prod_code, B1.batch_nbr) AS X
INNER JOIN
ProductTests AS B4
ON B4.prod_code = X.prod_code
AND B4.batch_nbr = X.batch_nbr
GROUP BY X.prod_code, X.end_batch_nbr;
クエリ解説1-1
元データ
select * from ProductTests;
batch_nbr | prod_code | prod_quality
-----------+-----------+--------------
1 | A | 80.0000
2 | A | 70.0000
3 | A | 80.0000
4 | B | 60.0000
5 | B | 90.0000
6 | C | 80.0000
7 | D | 80.0000
8 | A | 50.0000
9 | C | 70.0000
prod_codeが同じでbatch_nbrが大きいものと自己結合します
Aに限って考えると、Aのバッチ番号は{1,2,3,8}なので
{11,12,13,18, 22,23,28, 33,38, 88}の10通りのペアができます
※注意1 SELECT句の列指定は*としました。
※注意2 わかりやすさのためOrder byを加えています(行順意外に結果はありません)
batch_nbr | prod_code | prod_quality | batch_nbr | prod_code | prod_quality
-----------+-----------+--------------+-----------+-----------+--------------
1 | A | 80.0000 | 1 | A | 80.0000
1 | A | 80.0000 | 2 | A | 70.0000
1 | A | 80.0000 | 3 | A | 80.0000
1 | A | 80.0000 | 8 | A | 50.0000
2 | A | 70.0000 | 2 | A | 70.0000
2 | A | 70.0000 | 3 | A | 80.0000
2 | A | 70.0000 | 8 | A | 50.0000
3 | A | 80.0000 | 3 | A | 80.0000
3 | A | 80.0000 | 8 | A | 50.0000
4 | B | 60.0000 | 4 | B | 60.0000
4 | B | 60.0000 | 5 | B | 90.0000
5 | B | 90.0000 | 5 | B | 90.0000
6 | C | 80.0000 | 6 | C | 80.0000
6 | C | 80.0000 | 9 | C | 70.0000
7 | D | 80.0000 | 7 | D | 80.0000
8 | A | 50.0000 | 8 | A | 50.0000
9 | C | 70.0000 | 9 | C | 70.0000
(17 rows)
SELECT B1.prod_code, B1.batch_nbr,
MAX(B2.batch_nbr) AS end_batch_nbr
FROM ProductTests AS B1, ProductTests AS B2
WHERE B1.batch_nbr <= B2.batch_nbr
AND B1.prod_code = B2.prod_code

8. クエリ解説2※実行SQLは赤字で示します
SELECT X.prod_code, MIN(X.batch_nbr) AS start_batch_nbr, end_batch_nbr,
AVG(B4.prod_quality) AS avg_prod_quality
FROM (SELECT B1.prod_code, B1.batch_nbr,
MAX(B2.batch_nbr) AS end_batch_nbr
FROM ProductTests AS B1, ProductTests AS B2
WHERE B1.batch_nbr <= B2.batch_nbr
AND B1.prod_code = B2.prod_code
AND B1.prod_code
= ALL (SELECT prod_code
FROM ProductTests AS B3
WHERE B3.batch_nbr BETWEEN B1.batch_nbr
AND B2.batch_nbr)
GROUP BY B1.prod_code, B1.batch_nbr) AS X
INNER JOIN
ProductTests AS B4
ON B4.prod_code = X.prod_code
AND B4.batch_nbr = X.batch_nbr
GROUP BY X.prod_code, X.end_batch_nbr;
※注意1 SELECT句の列指定は*としました。
※注意2 わかりやすさのためOrder byを加えています(行順意外に結果はありません)
batch_nbr | prod_code | prod_quality | batch_nbr | prod_code | prod_quality
-----------+-----------+--------------+-----------+-----------+--------------
1 | A | 80.0000 | 1 | A | 80.0000
1 | A | 80.0000 | 2 | A | 70.0000
1 | A | 80.0000 | 3 | A | 80.0000
2 | A | 70.0000 | 2 | A | 70.0000
2 | A | 70.0000 | 3 | A | 80.0000
3 | A | 80.0000 | 3 | A | 80.0000
4 | B | 60.0000 | 4 | B | 60.0000
4 | B | 60.0000 | 5 | B | 90.0000
5 | B | 90.0000 | 5 | B | 90.0000
6 | C | 80.0000 | 6 | C | 80.0000
7 | D | 80.0000 | 7 | D | 80.0000
8 | A | 50.0000 | 8 | A | 50.0000
9 | C | 70.0000 | 9 | C | 70.0000
(13 rows)

9. ※実行SQLは赤字で示します
SELECT X.prod_code, MIN(X.batch_nbr) AS start_batch_nbr, end_batch_nbr,
AVG(B4.prod_quality) AS avg_prod_quality
FROM (SELECT B1.prod_code, B1.batch_nbr,
MAX(B2.batch_nbr) AS end_batch_nbr
FROM ProductTests AS B1, ProductTests AS B2
WHERE B1.batch_nbr <= B2.batch_nbr
AND B1.prod_code = B2.prod_code
AND B1.prod_code
= ALL (SELECT prod_code
FROM ProductTests AS B3
WHERE B3.batch_nbr BETWEEN B1.batch_nbr
AND B2.batch_nbr)
GROUP BY B1.prod_code, B1.batch_nbr) AS X
INNER JOIN
ProductTests AS B4
ON B4.prod_code = X.prod_code
AND B4.batch_nbr = X.batch_nbr
GROUP BY X.prod_code, X.end_batch_nbr;
SELECT B1.prod_code, B1.batch_nbr,
MAX(B2.batch_nbr) AS end_batch_nbr
FROM ProductTests AS B1, ProductTests AS B2
WHERE B1.batch_nbr <= B2.batch_nbr
AND B1.prod_code = B2.prod_code
AND B1.prod_code
= ALL (SELECT prod_code
FROM ProductTests AS B3
WHERE B3.batch_nbr BETWEEN B1.batch_nbr
AND B2.batch_nbr)
クエリ解説2-1
青色の行が消えます
batch_nbr | prod_code | prod_quality | batch_nbr | prod_quality
-----------+-----------+--------------+-----------+--------------
1 | A | 80.0000 | 1 | 80.0000
1 | A | 80.0000 | 2 | 70.0000
1 | A | 80.0000 | 3 | 80.0000
1 | A | 80.0000 | 8 | 50.0000
2 | A | 70.0000 | 2 | 70.0000
2 | A | 70.0000 | 3 | 80.0000
2 | A | 70.0000 | 8 | 50.0000
3 | A | 80.0000 | 3 | 80.0000
3 | A | 80.0000 | 8 | 50.0000
4 | B | 60.0000 | 4 | 60.0000
4 | B | 60.0000 | 5 | 90.0000
5 | B | 90.0000 | 5 | 90.0000
6 | C | 80.0000 | 6 | 80.0000
6 | C | 80.0000 | 9 | 70.0000
7 | D | 80.0000 | 7 | 80.0000
8 | A | 50.0000 | 8 | 50.0000
9 | C | 70.0000 | 9 | 70.0000
(17 rows)
※注意1 SELECT句の列指定は*としました。
※注意2 わかりやすさのためOrder byを加えています(行順意外に結果はありません)
batch_nbr | prod_code | prod_quality | batch_nbr | prod_code | prod_quality
-----------+-----------+--------------+-----------+-----------+--------------
1 | A | 80.0000 | 1 | A | 80.0000
1 | A | 80.0000 | 2 | A | 70.0000
1 | A | 80.0000 | 3 | A | 80.0000
2 | A | 70.0000 | 2 | A | 70.0000
2 | A | 70.0000 | 3 | A | 80.0000
3 | A | 80.0000 | 3 | A | 80.0000
4 | B | 60.0000 | 4 | B | 60.0000
4 | B | 60.0000 | 5 | B | 90.0000
5 | B | 90.0000 | 5 | B | 90.0000
6 | C | 80.0000 | 6 | C | 80.0000
7 | D | 80.0000 | 7 | D | 80.0000
8 | A | 50.0000 | 8 | A | 50.0000
9 | C | 70.0000 | 9 | C | 70.0000
(13 rows)
またまたAに限ったケースで考えると…

10. ※実行SQLは赤字で示します
SELECT X.prod_code, MIN(X.batch_nbr) AS start_batch_nbr, end_batch_nbr,
AVG(B4.prod_quality) AS avg_prod_quality
FROM (SELECT B1.prod_code, B1.batch_nbr,
MAX(B2.batch_nbr) AS end_batch_nbr
FROM ProductTests AS B1, ProductTests AS B2
WHERE B1.batch_nbr <= B2.batch_nbr
AND B1.prod_code = B2.prod_code
AND B1.prod_code
= ALL (SELECT prod_code
FROM ProductTests AS B3
WHERE B3.batch_nbr BETWEEN B1.batch_nbr
AND B2.batch_nbr)
GROUP BY B1.prod_code, B1.batch_nbr) AS X
INNER JOIN
ProductTests AS B4
ON B4.prod_code = X.prod_code
AND B4.batch_nbr = X.batch_nbr
GROUP BY X.prod_code, X.end_batch_nbr;
SELECT B1.prod_code, B1.batch_nbr,
MAX(B2.batch_nbr) AS end_batch_nbr
FROM ProductTests AS B1, ProductTests AS B2
WHERE B1.batch_nbr <= B2.batch_nbr
AND B1.prod_code = B2.prod_code
AND B1.prod_code
= ALL (SELECT prod_code
FROM ProductTests AS B3
WHERE B3.batch_nbr BETWEEN B1.batch_nbr
AND B2.batch_nbr)
クエリ解説2-2
青色の行が消えます
batch_nbr | prod_code | prod_quality | batch_nbr | prod_quality
-----------+-----------+--------------+-----------+--------------
1 | A | 80.0000 | 1 | 80.0000
1 | A | 80.0000 | 2 | 70.0000
1 | A | 80.0000 | 3 | 80.0000
1 | A | 80.0000 | 8 | 50.0000
2 | A | 70.0000 | 2 | 70.0000
2 | A | 70.0000 | 3 | 80.0000
2 | A | 70.0000 | 8 | 50.0000
3 | A | 80.0000 | 3 | 80.0000
3 | A | 80.0000 | 8 | 50.0000
4 | B | 60.0000 | 4 | 60.0000
4 | B | 60.0000 | 5 | 90.0000
5 | B | 90.0000 | 5 | 90.0000
6 | C | 80.0000 | 6 | 80.0000
6 | C | 80.0000 | 9 | 70.0000
7 | D | 80.0000 | 7 | 80.0000
8 | A | 50.0000 | 8 | 50.0000
9 | C | 70.0000 | 9 | 70.0000
(17 rows)
※注意1 SELECT句の列指定は*としました。
※注意2 わかりやすさのためOrder byを加えています(行順意外に結果はありません)
batch_nbr | prod_code | prod_quality | batch_nbr | prod_code | prod_quality
-----------+-----------+--------------+-----------+-----------+--------------
1 | A | 80.0000 | 1 | A | 80.0000
1 | A | 80.0000 | 2 | A | 70.0000
1 | A | 80.0000 | 3 | A | 80.0000
2 | A | 70.0000 | 2 | A | 70.0000
2 | A | 70.0000 | 3 | A | 80.0000
3 | A | 80.0000 | 3 | A | 80.0000
4 | B | 60.0000 | 4 | B | 60.0000
4 | B | 60.0000 | 5 | B | 90.0000
5 | B | 90.0000 | 5 | B | 90.0000
6 | C | 80.0000 | 6 | C | 80.0000
7 | D | 80.0000 | 7 | D | 80.0000
8 | A | 50.0000 | 8 | A | 50.0000
9 | C | 70.0000 | 9 | C | 70.0000
(13 rows)
ALL述語の条件は若干わかりにくいが、
自己結合したテーブル
B1とB2において、
B1.batch_nbr B2.batch_nbr
間のprod_codeの値が
全て一致する行を残すための相関サブクエリ
1 3や2 3間は全てAなので残る
2 8間では、4にB、6にCが混じるので
ALL述語が成り立たない。＝消える

11. クエリ解説3※実行SQLは赤字で示します
SELECT X.prod_code, MIN(X.batch_nbr) AS start_batch_nbr, end_batch_nbr,
AVG(B4.prod_quality) AS avg_prod_quality
FROM (SELECT B1.prod_code, B1.batch_nbr,
MAX(B2.batch_nbr) AS end_batch_nbr
FROM ProductTests AS B1, ProductTests AS B2
WHERE B1.batch_nbr <= B2.batch_nbr
AND B1.prod_code = B2.prod_code
AND B1.prod_code
= ALL (SELECT prod_code
FROM ProductTests AS B3
WHERE B3.batch_nbr BETWEEN B1.batch_nbr
AND B2.batch_nbr)
GROUP BY B1.prod_code, B1.batch_nbr) AS X
INNER JOIN
ProductTests AS B4
ON B4.prod_code = X.prod_code
AND B4.batch_nbr = X.batch_nbr
GROUP BY X.prod_code, X.end_batch_nbr;
prod_code | batch_nbr | end_batch_nbr
-----------+-----------+---------------
A | 1 | 3
A | 2 | 3
A | 3 | 3
B | 4 | 5
B | 5 | 5
C | 6 | 6
D | 7 | 7
A | 8 | 8
C | 9 | 9
(9 rows)

12. クエリ解説3-1※実行SQLは赤字で示します
SELECT X.prod_code, MIN(X.batch_nbr) AS start_batch_nbr, end_batch_nbr,
AVG(B4.prod_quality) AS avg_prod_quality
FROM (SELECT B1.prod_code, B1.batch_nbr,
MAX(B2.batch_nbr) AS end_batch_nbr
FROM ProductTests AS B1, ProductTests AS B2
WHERE B1.batch_nbr <= B2.batch_nbr
AND B1.prod_code = B2.prod_code
AND B1.prod_code
= ALL (SELECT prod_code
FROM ProductTests AS B3
WHERE B3.batch_nbr BETWEEN B1.batch_nbr
AND B2.batch_nbr)
GROUP BY B1.prod_code, B1.batch_nbr) AS X
INNER JOIN
ProductTests AS B4
ON B4.prod_code = X.prod_code
AND B4.batch_nbr = X.batch_nbr
GROUP BY X.prod_code, X.end_batch_nbr;
prod_code | batch_nbr | end_batch_nbr
-----------+-----------+---------------
A | 1 | 3
A | 2 | 3
A | 3 | 3
B | 4 | 5
B | 5 | 5
C | 6 | 6
D | 7 | 7
A | 8 | 8
C | 9 | 9
(9 rows)
対象データ
select * from ProductTests;
batch_nbr | prod_code | prod_quality
-----------+-----------+--------------
1 | A | 80.0000
2 | A | 70.0000
3 | A | 80.0000
4 | B | 60.0000
5 | B | 90.0000
6 | C | 80.0000
7 | D | 80.0000
8 | A | 50.0000
9 | C | 70.0000
(9 rows)
連続している製品グループ群の最後のbatch_nbrを
end_batch_nbrに格納している

13. クエリ解説4※実行SQLは赤字で示します
SELECT X.prod_code, MIN(X.batch_nbr) AS start_batch_nbr, end_batch_nbr,
AVG(B4.prod_quality) AS avg_prod_quality
FROM (SELECT B1.prod_code, B1.batch_nbr,
MAX(B2.batch_nbr) AS end_batch_nbr
FROM ProductTests AS B1, ProductTests AS B2
WHERE B1.batch_nbr <= B2.batch_nbr
AND B1.prod_code = B2.prod_code
AND B1.prod_code
= ALL (SELECT prod_code
FROM ProductTests AS B3
WHERE B3.batch_nbr BETWEEN B1.batch_nbr
AND B2.batch_nbr)
GROUP BY B1.prod_code, B1.batch_nbr) AS X
INNER JOIN
ProductTests AS B4
ON B4.prod_code = X.prod_code
AND B4.batch_nbr = X.batch_nbr
GROUP BY X.prod_code, X.end_batch_nbr;
※赤字の部分のまま実行できないのでselect * from ()で括ってサブクエリとして実行しています
prod_code | batch_nbr | end_batch_nbr | prod_quality
-----------+-----------+---------------+--------------
A | 1 | 3 | 80.0000
A | 2 | 3 | 70.0000
A | 3 | 3 | 80.0000
B | 4 | 5 | 60.0000
B | 5 | 5 | 90.0000
C | 6 | 6 | 80.0000
D | 7 | 7 | 80.0000
A | 8 | 8 | 50.0000
C | 9 | 9 | 70.0000
(9 rows)
内部結合でprod_qualityを付け足しただけ。

14. クエリ解説5※実行SQLは赤字で示します
SELECT X.prod_code, MIN(X.batch_nbr) AS start_batch_nbr, end_batch_nbr,
AVG(B4.prod_quality) AS avg_prod_quality
FROM (SELECT B1.prod_code, B1.batch_nbr,
MAX(B2.batch_nbr) AS end_batch_nbr
FROM ProductTests AS B1, ProductTests AS B2
WHERE B1.batch_nbr <= B2.batch_nbr
AND B1.prod_code = B2.prod_code
AND B1.prod_code
= ALL (SELECT prod_code
FROM ProductTests AS B3
WHERE B3.batch_nbr BETWEEN B1.batch_nbr
AND B2.batch_nbr)
GROUP BY B1.prod_code, B1.batch_nbr) AS X
INNER JOIN
ProductTests AS B4
ON B4.prod_code = X.prod_code
AND B4.batch_nbr = X.batch_nbr
GROUP BY X.prod_code, X.end_batch_nbr;
prod_code | start_batch_nbr | end_batch_nbr | avg_prod_quality
-----------+-----------------+---------------+---------------------
A | 1 | 3 | 76.6666666666666667
B | 4 | 5 | 75.0000000000000000
C | 6 | 6 | 80.0000000000000000
D | 7 | 7 | 80.0000000000000000
A | 8 | 8 | 50.0000000000000000
C | 9 | 9 | 70.0000000000000000
(6 rows)
赤字の列でgroup by。
・最小値(同じ製品のシーケンシャルなグループの開始位置)
・平均値
を集計関数で計算して表示

15. 完成
対象データ
select * from ProductTests;
batch_nbr | prod_code | prod_quality
-----------+-----------+--------------
1 | A | 80.0000
2 | A | 70.0000
3 | A | 80.0000
4 | B | 60.0000
5 | B | 90.0000
6 | C | 80.0000
7 | D | 80.0000
8 | A | 50.0000
9 | C | 70.0000
(9 rows)
prod_code | start_batch_nbr | end_batch_nbr | avg_prod_quality
-----------+-----------------+---------------+---------------------
A | 1 | 3 | 76.6666666666666667
B | 4 | 5 | 75.0000000000000000
C | 6 | 6 | 80.0000000000000000
D | 7 | 7 | 80.0000000000000000
A | 8 | 8 | 50.0000000000000000
C | 9 | 9 | 70.0000000000000000
(6 rows)

16. おさらい1※実行SQLは赤字で示します
SELECT X.prod_code, MIN(X.batch_nbr) AS start_batch_nbr, end_batch_nbr,
AVG(B4.prod_quality) AS avg_prod_quality
FROM (SELECT B1.prod_code, B1.batch_nbr,
MAX(B2.batch_nbr) AS end_batch_nbr
FROM ProductTests AS B1, ProductTests AS B2
WHERE B1.batch_nbr <= B2.batch_nbr
AND B1.prod_code = B2.prod_code
AND B1.prod_code
= ALL (SELECT prod_code
FROM ProductTests AS B3
WHERE B3.batch_nbr BETWEEN B1.batch_nbr
AND B2.batch_nbr)
GROUP BY B1.prod_code, B1.batch_nbr) AS X
INNER JOIN
ProductTests AS B4
ON B4.prod_code = X.prod_code
AND B4.batch_nbr = X.batch_nbr
GROUP BY X.prod_code, X.end_batch_nbr;
※注意1 SELECT句の列指定は*としました。
※注意2 わかりやすさのためOrder byを加えています(行順意外に結果はありません)
batch_nbr | prod_code | prod_quality | batch_nbr | prod_code | prod_quality
-----------+-----------+--------------+-----------+-----------+--------------
1 | A | 80.0000 | 1 | A | 80.0000
1 | A | 80.0000 | 2 | A | 70.0000
1 | A | 80.0000 | 3 | A | 80.0000
1 | A | 80.0000 | 8 | A | 50.0000
2 | A | 70.0000 | 2 | A | 70.0000
2 | A | 70.0000 | 3 | A | 80.0000
2 | A | 70.0000 | 8 | A | 50.0000
3 | A | 80.0000 | 3 | A | 80.0000
3 | A | 80.0000 | 8 | A | 50.0000
4 | B | 60.0000 | 4 | B | 60.0000
4 | B | 60.0000 | 5 | B | 90.0000
5 | B | 90.0000 | 5 | B | 90.0000
6 | C | 80.0000 | 6 | C | 80.0000
6 | C | 80.0000 | 9 | C | 70.0000
7 | D | 80.0000 | 7 | D | 80.0000
8 | A | 50.0000 | 8 | A | 50.0000
9 | C | 70.0000 | 9 | C | 70.0000
(17 rows)

17. おさらい2※実行SQLは赤字で示します
SELECT X.prod_code, MIN(X.batch_nbr) AS start_batch_nbr, end_batch_nbr,
AVG(B4.prod_quality) AS avg_prod_quality
FROM (SELECT B1.prod_code, B1.batch_nbr,
MAX(B2.batch_nbr) AS end_batch_nbr
FROM ProductTests AS B1, ProductTests AS B2
WHERE B1.batch_nbr <= B2.batch_nbr
AND B1.prod_code = B2.prod_code
AND B1.prod_code
= ALL (SELECT prod_code
FROM ProductTests AS B3
WHERE B3.batch_nbr BETWEEN B1.batch_nbr
AND B2.batch_nbr)
GROUP BY B1.prod_code, B1.batch_nbr) AS X
INNER JOIN
ProductTests AS B4
ON B4.prod_code = X.prod_code
AND B4.batch_nbr = X.batch_nbr
GROUP BY X.prod_code, X.end_batch_nbr;
※注意1 SELECT句の列指定は*としました。
※注意2 わかりやすさのためOrder byを加えています(行順意外に結果はありません)
batch_nbr | prod_code | prod_quality | batch_nbr | prod_code | prod_quality
-----------+-----------+--------------+-----------+-----------+--------------
1 | A | 80.0000 | 1 | A | 80.0000
1 | A | 80.0000 | 2 | A | 70.0000
1 | A | 80.0000 | 3 | A | 80.0000
2 | A | 70.0000 | 2 | A | 70.0000
2 | A | 70.0000 | 3 | A | 80.0000
3 | A | 80.0000 | 3 | A | 80.0000
4 | B | 60.0000 | 4 | B | 60.0000
4 | B | 60.0000 | 5 | B | 90.0000
5 | B | 90.0000 | 5 | B | 90.0000
6 | C | 80.0000 | 6 | C | 80.0000
7 | D | 80.0000 | 7 | D | 80.0000
8 | A | 50.0000 | 8 | A | 50.0000
9 | C | 70.0000 | 9 | C | 70.0000
(13 rows)

18. おさらい3※実行SQLは赤字で示します
SELECT X.prod_code, MIN(X.batch_nbr) AS start_batch_nbr, end_batch_nbr,
AVG(B4.prod_quality) AS avg_prod_quality
FROM (SELECT B1.prod_code, B1.batch_nbr,
MAX(B2.batch_nbr) AS end_batch_nbr
FROM ProductTests AS B1, ProductTests AS B2
WHERE B1.batch_nbr <= B2.batch_nbr
AND B1.prod_code = B2.prod_code
AND B1.prod_code
= ALL (SELECT prod_code
FROM ProductTests AS B3
WHERE B3.batch_nbr BETWEEN B1.batch_nbr
AND B2.batch_nbr)
GROUP BY B1.prod_code, B1.batch_nbr) AS X
INNER JOIN
ProductTests AS B4
ON B4.prod_code = X.prod_code
AND B4.batch_nbr = X.batch_nbr
GROUP BY X.prod_code, X.end_batch_nbr;
prod_code | batch_nbr | end_batch_nbr
-----------+-----------+---------------
A | 1 | 3
A | 2 | 3
A | 3 | 3
B | 4 | 5
B | 5 | 5
C | 6 | 6
D | 7 | 7
A | 8 | 8
C | 9 | 9
(9 rows)

19. おさらい4※実行SQLは赤字で示します
SELECT X.prod_code, MIN(X.batch_nbr) AS start_batch_nbr, end_batch_nbr,
AVG(B4.prod_quality) AS avg_prod_quality
FROM (SELECT B1.prod_code, B1.batch_nbr,
MAX(B2.batch_nbr) AS end_batch_nbr
FROM ProductTests AS B1, ProductTests AS B2
WHERE B1.batch_nbr <= B2.batch_nbr
AND B1.prod_code = B2.prod_code
AND B1.prod_code
= ALL (SELECT prod_code
FROM ProductTests AS B3
WHERE B3.batch_nbr BETWEEN B1.batch_nbr
AND B2.batch_nbr)
GROUP BY B1.prod_code, B1.batch_nbr) AS X
INNER JOIN
ProductTests AS B4
ON B4.prod_code = X.prod_code
AND B4.batch_nbr = X.batch_nbr
GROUP BY X.prod_code, X.end_batch_nbr;
※赤字の部分のまま実行できないのでselect * from ()で括ってサブクエリとして実行しています
prod_code | batch_nbr | end_batch_nbr | prod_quality
-----------+-----------+---------------+--------------
A | 1 | 3 | 80.0000
A | 2 | 3 | 70.0000
A | 3 | 3 | 80.0000
B | 4 | 5 | 60.0000
B | 5 | 5 | 90.0000
C | 6 | 6 | 80.0000
D | 7 | 7 | 80.0000
A | 8 | 8 | 50.0000
C | 9 | 9 | 70.0000
(9 rows)
内部結合でprod_qualityを付け足しただけ。

20. おさらい5※実行SQLは赤字で示します
SELECT X.prod_code, MIN(X.batch_nbr) AS start_batch_nbr, end_batch_nbr,
AVG(B4.prod_quality) AS avg_prod_quality
FROM (SELECT B1.prod_code, B1.batch_nbr,
MAX(B2.batch_nbr) AS end_batch_nbr
FROM ProductTests AS B1, ProductTests AS B2
WHERE B1.batch_nbr <= B2.batch_nbr
AND B1.prod_code = B2.prod_code
AND B1.prod_code
= ALL (SELECT prod_code
FROM ProductTests AS B3
WHERE B3.batch_nbr BETWEEN B1.batch_nbr
AND B2.batch_nbr)
GROUP BY B1.prod_code, B1.batch_nbr) AS X
INNER JOIN
ProductTests AS B4
ON B4.prod_code = X.prod_code
AND B4.batch_nbr = X.batch_nbr
GROUP BY X.prod_code, X.end_batch_nbr;
prod_code | start_batch_nbr | end_batch_nbr | avg_prod_quality
-----------+-----------------+---------------+---------------------
A | 1 | 3 | 76.6666666666666667
B | 4 | 5 | 75.0000000000000000
C | 6 | 6 | 80.0000000000000000
D | 7 | 7 | 80.0000000000000000
A | 8 | 8 | 50.0000000000000000
C | 9 | 9 | 70.0000000000000000
(6 rows)
赤字の列でgroup by。
・最小値(同じ製品のシーケンシャルなグループの開始位置)
・平均値
を集計関数で計算して表示