在SQL中没有用于组合的数据时创建空行

如何解决在SQL中没有用于组合的数据时创建空行

大多数人可能已经问过这个问题，我也曾尝试过，但无法解决。我有一个表，其中包含15个具有非度量的属性/列和15个具有度量的列。对于非度量列-如果不存在任何组合，我想创建空值。我正在使用可以具有简单PROC SQL语句的SAS。可以说我有如下数据： enter image description here

Name | Location | Holiday | Time | DishCnt | TastedGood
Bill    1st Fl.    Easter    AM       3          Yes
Bill    1st Fl.    Easter    PM       3          Yes
Bill    1st Fl.    Kawanza   AM       2          Yes
Bill    2st Fl.    Kawanza   AM       2          Yes
Sameer  1st Fl.    Easter    AM       1          No
Sameer  1st Fl.    Easter    PM       2          Yes
Sameer  1st Fl.    Kawanza   AM       1          Yes
Peter   1st Fl.    Easter    AM       1          No
Peter   1st Fl.    Kawanza   AM       2          No

我想检查每个非测量行的组合，如果没有，那么我想创建一个新行并将其设置为null或0。 enter image description here 我想要的样本数据如下：

Name | Location | Holiday | Time | DishCnt | TastedGood
Bill    1st Fl.    Easter    AM       3          Yes
Bill    1st Fl.    Easter    PM       3          Yes
Bill    1st Fl.    Kawanza   AM       2          Yes
Bill    2st Fl.    Kawanza   AM       2          Yes
Sameer  1st Fl.    Easter    AM       1          No
Sameer  1st Fl.    Easter    PM       2          Yes
Sameer  1st Fl.    Kawanza   AM       1          Yes
Sameer  2st Fl.    Kawanza   AM      NULL        NULL
Peter   1st Fl.    Easter    AM       1          No
Peter   1st Fl.    Kawanza   AM       2          No
Peter   1st Fl.    Easter    PM      NULL        NULL
Peter   2st Fl.    Kawanza   AM      NULL        NULL

我做的代码是我试图交叉连接同一张表并与公共列连接。像这样的东西

SELECT * 
FROM Holidays H
CROSS JOIN Holidays T
LEFT JOIN Holidays R ON H.HoludayKey = R.HolidayKey AND T.TimeKey=R.TimeKey

预先感谢您的帮助。

谢谢！

解决方法

请考虑交叉连接的数据集，以获取名称，位置，假日和时间的所有可能值。然后使用LEFT JOIN运行此查询到实际数据集。任何不匹配的行都将填充NULL或丢失。

proc sql;
   CREATE TABLE combns AS
   SELECT n.Name,m.Location,m.Holiday,m.Time
   FROM 
       (SELECT DISTINCT Name
        FROM Holidays) n
   CROSS JOIN
       (SELECT DISTINCT Location,Holiday,Time
        FROM Holidays) m;
   
   CREATE TABLE final_table AS
   SELECT c.Name,c.Location,c.Holiday,c.Time,h.DishCnt,h.TastedGood
   FROM combns c
   LEFT JOIN Holidays h
    ON c.Name = h.Name 
    AND c.Location = h.Location
    AND c.Holiday = h.Holiday
    AND c.Time = h.Time;
quit;

---

输入 （OP的发布数据）

data Holidays;
    infile datalines delimiter=',' DSD; 
    length Name $8 Location $8 Holiday $8 Time $2 DishCnt TastedGood $3;
    
    input Name Location Holiday Time DishCnt TastedGood;
    datalines;
Bill,1st Fl.,Easter,AM,3,Yes
Bill,PM,Kawanza,2,2st Fl.,Yes
Sameer,1,No
Sameer,Yes
Peter,No
Peter,No
;

输出 （在proc sql上方运行之后）