Question

我对Proc sql很新，觉得我正撞在墙上。基本上，我想制作一些不同类型的小计，我一直在努力。我可以弄清楚如何在两个sql语句中执行它，但我无法弄清楚如何使用子查询将它们组合成一个语句，这似乎应该是可能的。

示例数据

*job, gender
*florist, female
*florist, female
*florist, male
*florist, male
*Manager, female
*Manager, female
*Manager, female
*Manager, male
*nurse, female
*nurse, female
*nurse, male

我想从我的代码中得到的是：

Job Title|  Gender  | All answers to gender | Job by gender
Florist  |  female  | 7                     | 2
Florist  |  male    | 4                     | 2
Manager  |  female  | 7                     | 3
Manager  |  male    | 4                     | 1
nurse    |  female  | 7                     | 2
nurse    |  male    | 4                     | 1

我当前的代码（分两步执行）是

proc sql;
    create table newdata as
          select gender, job,
            count(gender) as GenderCount
          from datasource where gender in ('f','m')
          group by gender;

     select * from newdata;
quit

proc sql;
     create table newdata2 as 
            select *,
             count(gender) as JobsByGender
            from newdata
            group by q1, q21;

      select distinct * from newdata2 order by q21, q1;
      quit;

Answer 1

您可以加入子查询以获得预期结果：

proc sql;
    create table newdata as
        select bd.job,
            bd.gender,
            sq.count_gender,
            count(bd.job) as JobsByGender
        from basedata bd
        inner join
            (select gender,
                 count(gender) as count_gender
             from base_data
             group by gender) sq
        on sq.gender = bd.gender
        group by bd.job, bd.gender;
quit;

子查询创建一个看起来像

的数据集

Gender | count_gender
---------------------
Male   | 4
Female | 7

内部联接会根据on sq.gender = bd.gender

指定的每个性别匹配正确的数字

老实说，大部分只是通用SQL。如果您将来要做更多的proc sql语句，我强烈建议您查看可以在http://www.sqlcourse.com/和http://www.sqlcourse2.com/找到的教程

这些希望能让您理解您正在撰写的查询，而不是使用您不理解的示例代码。

Proc sql小计和子查询

1 个答案: