Question

我是SQL的新手，我正尝试自己学习T-SQL，这是一本初学者的书。我陷入了一个查询，仍然无法解决；为什么起作用，或如何起作用。我希望你能帮助我。

我有桌子：

CREATE TABLE #AnnualSales
(
    YEAR CHAR(4),
    SALE INT
);

INSERT INTO #AnnualSales (YEAR, SALE)
VALUES ('2001', 100), ('2002', 150), ('2003', 120), ('2004', 150),
       ('2005', 250), ('2006', 300), ('2007', 250);

目标是确定按年份排序的销售趋势，其中“ 1”为正，“-1”为负。这本书说我们可以从（1）中做到这一点：

WITH Years AS
(
    SELECT 
        Z.Year, Z.Sale, 
        SIGN(Z.Sale - (SELECT W1.Sale
                       FROM #AnnualSales AS W1
                       WHERE W1.Year = (SELECT MAX(W2.Year)
                                        FROM #AnnualSales AS W2
                                        WHERE W2.Year < Z.Year))) AS Trend
    FROM 
        #AnnualSales AS Z
)
SELECT 
    *, 
    (SELECT COUNT(*)
     FROM Years AS T2
     WHERE T2.Year >= T1.Year 
       AND T2.Trend <> T1.Trend) AS Rows
FROM 
    Years AS T1

对此（2）：

WITH Years AS
(
    SELECT 
        Z.Year, Z.Sale, 
        SIGN(Z.Sale - (SELECT W1.Sale
                       FROM #AnnualSales AS W1
                       WHERE W1.Year = (SELECT MAX(W2.Year)
                                        FROM #AnnualSales AS W2
                                        WHERE W2.Year<Z.Year))) AS Trend
    FROM 
        #AnnualSales AS Z
),
Factor AS
(
    SELECT 
        *, 
        (SELECT COUNT(*)
         FROM Years AS T2
         WHERE T2.Year >= T1.Year AND T2.Trend <> T1.Trend) AS Rows
    FROM 
        Years AS T1
)
SELECT 
    MIN(Year), MAX(Year), Trend
FROM 
    Factor
GROUP BY 
    Rows, Trend
ORDER BY 
    MIN(Year)

我不知道它从（1）到（2）的工作方式。我不明白为什么查询（1）中的结果显示诸如'2'和'3'之类的值，最后一切都按预期运行。

通过SQL查询确定趋势；代码说明

0 个答案: