根据一个变量的值将 SAS 数据集拆分为多个数据集 [英] Splitting a SAS dataset into multiple datasets, according to value of one variable

问题描述

是否有更简化的方法来实现这一点?这是一个简化的例子.在实际情况下，var 有 > 10 个值，每个值都需要自己的数据集.

数据新1新2新3;设置旧;如果 var = 'new1' 然后输出 new1;否则如果 var = 'new2' 然后输出 new2;else if var = 'new3' 然后输出 new3;跑步;

解决方案

这应该可以解决.您只需要将 % 更改为 5 到 10(最大新数字).@Reeza 提出的观点很棒.我也会看看那个帖子，因为它是一个重要的建议.通常这不是处理数据的好方法，但这应该可以帮助您解决问题.

数据有；输入变量 $;数据线；新1新2新3新4新5;跑步;

<块引用>

*实际代码从这里开始；

%宏拆分；%do i=1 % 到 5;%put "new&i";进程sql;创建表 table&i as选择 *从有其中 var 包含new&i"；放弃;％结尾;%修正分裂；%分裂;

Is there a more streamlined way of accomplishing this? This is a simplified example. In the real case there are > 10 values of var, each of which need their own dataset.

data 
  new1
  new2
  new3;
set old;

if var = 'new1' then output new1;
else if var = 'new2' then output new2;
else if var = 'new3' then output new3;
run;

解决方案

This should work out. You just need to change the %to 5 to 10 (the max new number). The point made by @Reeza is great. I would also take a look at that post since it's an important suggestion. Usually this is not a good way to handle data, but this should get you around.

data have;
input      var $;
datalines;
            new1
            new2 
            new3 
            new4 
            new5 
;
run;

*Actual code starts here;

%macro splitting;
%do i=1 %to 5;

%put "new&i";

proc sql;
    create table table&i as 
    select *
    from have
    where var contains "new&i";
quit;

%end;
%mend splitting; 

%splitting;

这篇关于根据一个变量的值将 SAS 数据集拆分为多个数据集的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！