Question

我在vfp中使用SQL，所以我的一些命令是不同的。使用iif([condition],[if true],[if false])代替case when。

这是我的代码：

SELECT
    giftsource,
    COUNT(donor) AS total_gifts, /*this is not the one that need fixed*/
    SUM(amt) AS total_amt,
    SUM(iif(unique = 1 AND language != 'F' AND renew = '0' AND type = 'IN',1,0)) as d_Count_E_New_Indiv,
    SUM(iif(unique = 1 AND language = 'F' AND renew = '0' AND type = 'IN',1,0)) as d_Count_F_New_Indiv,
    SUM(iif(unique = 1 AND language != 'F' AND renew != '0' AND type = 'IN',1,0)) as d_Count_E_Re_Indiv,
    SUM(iif(unique = 1 AND language = 'F' AND renew != '0' AND type = 'IN',1,0)) as d_Count_F_Re_Indiv,
    SUM(iif(unique = 1 AND language != 'F' AND renew = '0' AND type != 'IN',1,0)) as d_Count_E_New_Org,
    SUM(iif(unique = 1 AND language = 'F' AND renew = '0' AND type != 'IN',1,0)) as d_Count_F_New_Org,
    SUM(iif(unique = 1 AND language != 'F' AND renew != '0' AND type != 'IN',1,0)) as d_Count_E_Re_Org,
    SUM(iif(unique = 1 AND language = 'F' AND renew != '0' AND type != 'IN',1,0)) as d_Count_F_Re_Org,
FROM (select *,
            cast( /* equivalent to a bunch of if elses*/
                iif(  list_code="WEB","1",
                iif(  list_code="GRO","2",
                iif(  list_code="CHO","3",
                iif(  list_code="TEL","4",
                iif(  list_code="TES","5",
                iif(  list_code="POS" AND amt < 10000,"6",
                iif( (LIKE(list_code,"4%") OR list_code = "4") AND amt < 10000,"7",
                iif( (LIKE(list_code,"4%") OR list_code = "4" OR list_code = "POS") AND amt >= 10000,"8",
                iif(  LIKE(list_code,"9%") OR list_code = "9","9",
                "10"))))))))) as c(1))
            as giftsource
        from cGift) gift
    LEFT JOIN
            (select didnumb, language, type from dp) d
        on cast(gift.donor as i) = cast(d.didnumb as i)
    LEFT JOIN /*this does not do what i want it to*/
            (select min(gidnumb) gid, 1 as unique from cGift group by donor) uGift
        on gift.gidnumb = uGift.gid
GROUP BY giftsource
ORDER BY giftsource

此代码应该做的是查找每个礼品类别中捐赠的捐赠者数量。它不应该计算相同字段内的重复捐赠者（即相同的list_code / lang / renew / etc.），但如果它是另一个字段，它应该计算两次捐赠者。

示例：捐赠者＃3只能在d_Count_E_New_Indiv中计算一次，但也可以在d_Count_E_New_Org中计算。

gidnumb是此表中的主键。

在我的第二次加入中，我将一个字段（名为unique）附加到表中的第一个捐赠者。这不起作用，因为它只计算一个领域的捐赠者。

有人可以告诉我，正确的做法是什么？此外，我有更多的SUM（...）不是基于独特的捐赠者，所以我宁愿不要过多地选择我的选择。

编辑：我使用以下select count(distinct IIF(renew = '0' AND lang != 'F',donor,0)) FROM dpgift

修复了它

Answer 1

我认为你可以逃脱COUNT（不同的捐赠者）AS total_gifts，我没有视觉狐狸专业人士，并做了一些谷歌搜索，并提到它。

Answer 2

如何使用在所有相关字段上分组的单个查询（看起来像giftource，语言，续订和类型），以便为每个组合获取一条记录，并显示所需的计数：

SELECT giftsource, language, renew, type, COUNT(*) ;
  FROM <whatever tables and joins you need> ;
  GROUP BY giftsource, language, renew, type

然后，您可以将此连接到您需要的任何其他内容，或使用一些Xbase代码将其拉入单行。您还可以查看交叉表（使用VFPXTab或第三方工具之一）来转移数据。

当字段是唯一的时，选择count（）

2 个答案: