ERROR: File XXXX is sequential. This task requires reading observations in a random order, but the engine allows only sequential access

Question

我正在尝试更新 oracle table。

我在 运行 以下数据步骤时遇到此错误：

data oracle.have(drop=_:);
    modify oracle.have end=last;
    if _n_=1 then do;
        declare hash h1(dataset:'_update');
        declare hiter hh1('h1');
        _rc = h1.defineKey('id','tid','valid_to');
        _rc = h1.defineData('valid_from');
        _rc = h1.defineDone();
    end;

    if h1.find()=0 then do;
        replace;
        _rc = h1.remove();
    end;
run;

ERROR: File ORACLE.HAVE.DATA is sequential. This task requires reading observations in a random order, but the engine allows only sequential access.

有什么方法可以绕过这个错误吗？

Answer 1

显然，修改语句仅适用于 SAS 数据集。

我使用的解决方法如下：

proc sql;
update have t1
       set valid_from = (select valid_from from _update t2
                         where t1.id = t2.id
                            and t1.tid = t2.tid
                            and t1.valid_to = t2.valid_to
                            and t2.dim="FROM")
      where catx('#',id, tid, valid_to) in (select catx('#',id, tid, valid_to)
                                            from _update t3
                                            where t3.dim="FROM");
quit;