sql - 使用连接条件单独从各个表中检索不同计数

Question

我需要查看 2013 年和 2014 年有多少客户在场

以下是我使用的查询:

proc sql;
        select count(distinct a.CUSTOMER_NUMBER)
        from LOCFIL.CUST_DB_2013 as a
        inner join LOCFIL.CUST_DB_2014 as b
            on a.CUSTOMER_NUMBER = B.CUSTOMER_NUMBER
;quit;

这给了我这两年的客户数量

但我需要 1) 仅限 2013 年的客户，并持续到 2014 年 2) 仅限 2014 年且 2013 年也出席过的客户

为此我尝试了以下方法:

1)

proc sql;
        select count(distinct a.CUSTOMER_NUMBER)
        from LOCFIL.CUST_DB_2013 as a
        left  join LOCFIL.CUST_DB_2014 as b
            on a.CUSTOMER_NUMBER = B.CUSTOMER_NUMBER
;quit;

2)

proc sql;
        select count(distinct a.CUSTOMER_NUMBER)
        from LOCFIL.CUST_DB_2014 as a
        left  join LOCFIL.CUST_DB_2013 as b
            on a.CUSTOMER_NUMBER = B.CUSTOMER_NUMBER
;quit;

但是数字甚至不接近(左连接查询计数比内部连接的计数多，这让我担心)。 您能帮助我如何获得这些号码吗？

Answer 1

如果有人需要在两年内都在场，请使用内部加入:

proc sql;
        select count(distinct a.CUSTOMER_NUMBER)
        from LOCFIL.CUST_DB_2013 as a join
             LOCFIL.CUST_DB_2014 as b
             on a.CUSTOMER_NUMBER = B.CUSTOMER_NUMBER;
quit;

如果每年没有重复的客户，则只需使用 count(*)。