分析得很清楚。

不错，分析的很细致

还是通过例子能够更好理解区别

merge Examples

Example 1: One-to-One Merging
This example shows how to combine observations from two data sets into a single observation in a new data set:

data benefits.qtr1;
   merge benefits.jan benefits.feb;
run;



Example 2: Match-Merging
This example shows how to combine observations from two data sets into a single observation in a new data set according to the values of a variable that is specified in the BY statement:

data inventry;
   merge stock orders;
   by partnum;
run;



Example 3: Merging with a Data Set List
This example uses a data list to define the data sets that are merged.

data d008; job=3; emp=19; run;
data d009; job=3; sal=50; run;
data d010; job=4; emp=97; run;
data d011; job=4; sal=15; run;
data comb;
merge d008-d011;
by job;
run;
proc print data=comb;
run;

update Examples

Example 1: Basic Updating
These program statements create a new data set (OHIO.QTR1) by applying transactions to a master data set (OHIO.JAN). The BY variable STORE must appear in both OHIO.JAN and OHIO.WEEK4, and its values in the master data set should be unique:

data ohio.qtr1;
   update ohio.jan ohio.week4;
   by store;
run;



Example 2: Updating By Renaming Variables
This example shows renaming a variable in the FITNESS data set so that it will not overwrite the value of the same variable in the program data vector. Also, the WEIGHT variable is renamed in each data set and a new WEIGHT variable is calculated. The master data set and the transaction data set are listed before the code that performs the update:

Master Data Set                          
            HEALTH                     

OBS    ID     NAME     TEAM    WEIGHT      

1    1114    sally    blue      125      
2    1441    sue      green     145      
3    1750    joey     red       189      
4    1994    mark     yellow    165
5    2304    joe      red       170

Transaction Data Set
            FITNESS

OBS   ID     NAME     TEAM    WEIGHT

1   1114    sally    blue      119
2   1994    mark     yellow    174
3   2304    joe      red       170


options nodate pageno=1 linesize=80 pagesize=60;

   /* Sort both data sets by ID */
proc sort data=health;
   by id;
run;
proc sort data=fitness;
   by id;
run;

   /* Update Master with Transaction */
data health2;
   length STATUS $11;
   update health(rename=(weight=ORIG) in=a)
          fitness(drop=name team in=b);
   by id ;
   if a and b then
      do;
         CHANGE=abs(orig - weight);
         if weight<orig then status='loss';
         else if weight>orig then status='gain';
         else status='same';
      end;
   else status='no weigh in';
run;

options nodate ls=78;

proc print data=health2;
   title 'Weekly Weigh-in Report';
run;

Updating By Renaming Variables

                     Weekly Weigh-in Report             1

  OBS STATUS        ID   NAME   TEAM    ORIG  WEIGHT  CHANGE

   1  loss         1114  sally  blue     125    119      6  
   2  no weigh in  1441  sue    green    145      .      .  
   3  no weigh in  1750  joey   red      189      .      .  
   4  gain         1994  mark   yellow   165    174      9  
   5  same         2304  joe    red      170    170      0  


Example 3: Updating with Missing Values
This example illustrates the DATA steps used to create a master data set PAYROLL and a transaction data set INCREASE that contains regular and special missing values:

options nodate pageno=1 linesize=80 pagesize=60;

   /* Create the Master Data Set */
data payroll;
   input ID SALARY;
   datalines;
011 245
026 269
028 374
034 333
057 582
;

   /* Create the Transaction Data Set */
data increase;
   input ID SALARY;
   missing A _;
   datalines;
011 376
026 .
028 374
034 A
057 _
;

   /* Update Master with Transaction */
data newpay;
   update payroll increase;
   by id;
run;
proc print data=newpay;
   title 'Updating with Missing Values';
run;

Updating With Missing Values

       Updating with Missing Values              1

           OBS     ID     SALARY

            1     1011      376
            2     1026      269    <=== value remains 269
            3     1028      374
            4     1034        A    <=== special missing value
            5     1057        .    <=== regular missing value