Question

假设我有以下数据：

表1（我的主数据集）：

clear 
input str1 Name str9 (Day Time)   
 A     24-Mar-08  "10:30:01"       
 A     24-Mar-08  "10:30:01"      
 B     24-Mar-08  "10:30:01"      
 B     24-Mar-08  "11:03:12"       
 B     24-Mar-08  "11:03:12"       
 C     25-Mar-08  "10:30:01"       
 D     25-Mar-08  "10:30:01"      
 D     25-Mar-08  "11:13:59"       
 E     25-Mar-08  "11:13:59"       
 F     25-Mar-08  "11:13:59"       
 F     25-Mar-08  "11:13:59"       
 F     25-Mar-08  "11:59:01"       
end

表2：

clear 
input str1 Insrument str9 (Day Time) float price  
 A     30-Mar-10  "09:29:34"  10    
 E     03-Mar-09  "08:23:19"  14   
 E     20-Mar-12  "12:15:11"  29    
 F     01-Mar-11  "10:30:01"  12     
end

我想根据表1中的name变量删除重复的观察结果。我在Stata中打开表1，需要将表2引入Stata，然后删除两个表中具有相同名称的相同观察，来自表1。

最后，结果是：

+------+------------+----------+
| name |    date    |    time  |
+------+------------+----------+
| B    |  24-Mar-08 | 10:30:01 |
| B    |  24-Mar-08 | 11:03:12 |
| B    |  24-Mar-08 | 11:03:12 |
| C    |  25-Mar-08 | 10:30:01 |
| D    |  25-Mar-08 | 10:30:01 |
| D    |  25-Mar-08 | 11:13:59 |
+------+------------+----------+

SAS中的SQL代码是：

proc SQL;
    create table table3 as
    select * from table1
    where name not in (select Instrument from table2);
quit;

如何使用Stata使用此程序？

Answer 1

最简单的方法是对名称进行合并，只保留那些没有匹配的名称：

clear 
input str1 Name str9 (Day Time)   
 A     24-Mar-08  "10:30:01"       
 E     24-Mar-08  "10:30:01"      
 E     24-Mar-08  "10:30:01"      
 F     24-Mar-08  "11:03:12"       
end 

keep Name 
duplicates drop
tempfile names
save "`names'"

clear 
input str1 Name str9 (Day Time)   
 A     24-Mar-08  "10:30:01"       
 A     24-Mar-08  "10:30:01"      
 B     24-Mar-08  "10:30:01"      
 B     24-Mar-08  "11:03:12"       
 B     24-Mar-08  "11:03:12"       
 C     25-Mar-08  "10:30:01"       
 D     25-Mar-08  "10:30:01"      
 D     25-Mar-08  "11:13:59"       
 E     25-Mar-08  "11:13:59"       
 F     25-Mar-08  "11:13:59"       
 F     25-Mar-08  "11:13:59"       
 F     25-Mar-08  "11:59:01"       
end 

merge m:1 Name using "`names'", keep(master) nogen
list, clean noobs

这会产生：

Name         Day       Time  
   B   24-Mar-08   10:30:01  
   B   24-Mar-08   11:03:12  
   B   24-Mar-08   11:03:12  
   C   25-Mar-08   10:30:01  
   D   25-Mar-08   10:30:01  
   D   25-Mar-08   11:13:59

Answer 2

// Keeping equities
rename name Instrument
merge m:1 Instrument using "file name path"
keep if _merge == 3
keep Instrument Day Time

在两个表

2 个答案: