以下是引用eblog在2008-8-24 9:53:00的发言:
再配上试用数据就更完美了
* 处理命令
*- 伪造一份非平行的面板数据
xtarsim y x eta, n(200) t(10) g(0) b(.8) r(.2) sn(9) seed(1234)
sum
replace tvar = tvar+1990
drop if x > 8                  /* unblanced */
replace y = . if abs(y) < 0.08  /* y 包含缺漏值 */
replace x = . if abs(x) < 0.03  /* x 包含缺漏值 */
tsset ivar tvar
xtdes             /*转换前的数据结果*/
*- 使用 xtbalance 命令进行转换
xtbalance, range(1991 2000) miss(_all) 
* ==========结果===================
. clear
. ssc install xtarsim
checking xtarsim consistency and verifying not already installed...
all files already exist and are up to date.
. 
. *- 伪造一份非平行的面板数据
. xtarsim y x eta, n(200) t(10) g(0) b(.8) r(.2) sn(9) seed(1234)
. sum
    Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
        ivar |      2000       100.5    57.74874          1        200
        tvar |      2000         5.5       2.873          1         10
           y |      2000   -.2178288    3.324586  -10.44931   10.21746
           x |      2000    -.143428    3.783924  -13.00195   11.69052
         eta |      2000   -.1048423    .9430213  -3.450946   2.096194
. replace tvar = tvar+1990
(2000 real changes made)
. drop if x > 8                  /* unblanced */
(27 observations deleted)
. replace y = . if abs(y) < 0.08  /* y 包含缺漏值 */
(36 real changes made, 36 to missing)
. replace x = . if abs(x) < 0.03  /* x 包含缺漏值 */
(8 real changes made, 8 to missing)
. tsset ivar tvar
       panel variable:  ivar (unbalanced)
        time variable:  tvar, 1991 to 2000, but with gaps
. xtdes             /*转换前的数据结果*/
    ivar:  1, 2, ..., 200                                    n =        200
    tvar:  1991, 1992, ..., 2000                             T =         10
           Delta(tvar) = 1; (2000-1991)+1 = 10
           (ivar*tvar uniquely identifies each observation)
Distribution of T_i:   min      5%     25%       50%       75%     95%     max
                         8       9      10        10        10      10      10
     Freq.  Percent    Cum. |  Pattern
 ---------------------------+------------
      176     88.00   88.00 |  1111111111
        6      3.00   91.00 |  111111111.
        3      1.50   92.50 |  .111111111
        3      1.50   94.00 |  1111.11111
        3      1.50   95.50 |  111111.111
        2      1.00   96.50 |  1.11111111
        2      1.00   97.50 |  111.111111
        1      0.50   98.00 |  1..1111111
        1      0.50   98.50 |  11..111111
        3      1.50  100.00 | (other patterns)
 ---------------------------+------------
      200    100.00         |  XXXXXXXXXX
. 
. *- 使用 xtbalance 命令进行转换
. xtbalance, range(1991 2000) miss(_all) 
(44 observations deleted due to missing) 
(519 observations deleted due to discontinues) 
. 
. xtdes             /*转换后的数据结果*/ 
    ivar:  1, 2, ..., 200                                    n =        141
    tvar:  1991, 1992, ..., 2000                             T =         10
           Delta(tvar) = 1; (2000-1991)+1 = 10
           (ivar*tvar uniquely identifies each observation)
Distribution of T_i:   min      5%     25%       50%       75%     95%     max
                        10      10      10        10        10      10      10
     Freq.  Percent    Cum. |  Pattern
 ---------------------------+------------
      141    100.00  100.00 |  1111111111
 ---------------------------+------------
      141    100.00         |  XXXXXXXXXX

[em05] 
[此贴子已经被作者于2009-3-29 10:39:14编辑过]
eblog  金币 +3  金钱 +25  魅力 +15  经验 +25  好文章 2008-8-28 13:59:48