这个版本可以估计Berry, Levinsohn, and Pakes模型，目前是IO里面比较时髦的。
程序：
SAMPLE ; 1-1000 $
CREATE ; market = Trn(10,0) $ (There will be 10 brands in each market) $
SETPANEL ; Group = market ; Pds = mkt $
CREATE ; x1 = Rnn(0,.2) ; x2 = Rnn(0,.1) ; x3 =Rnn(0,.1) $ (Attributes and price) $
CREATE ; u = Rnn(1,.05)*x2+x1+Rnn(-.5,.3)*x3+.6 $ (Utilities with RPs) $
CREATE ; eu = Exp(u) $
CREATE ; iv = Group Sums(eu, Pds = mkt) $ (Inclusive value) $
CREATE ; iv = iv+1 $ (Includes base opt-out brand) $
CREATE ; eu = Rnu(.7,.9)*eu/iv $
CREATE ; total = Group Sums(eu, Pds = mkt) $
CREATE ; shares = eu $ (Random market shares) $
? Instruments are 2 exogenous variables and products of exogenous attributes.
CREATE ; z1 = Rnu(1,1.5) ; z2 = x1 ; z3 = x2 ; z4 = x1*x2 ; z5 = x1*x1 ; z6 = x2*x2 $
CREATE ; z7 = Rnn(0,1) ; z8 = z7*z7 $
NAMELIST ; z = one,z* $ (All instruments) $
NAMELIST ; x = one,x1,x2,x3 $ (All attributes) $
BLPLOGIT ; Lhs = shares ; Rhs = one,x1 ; RPL = x2,x3
; Inst = z
; Draws = 50
; markets = 10 $
结果：
Iterative procedure has converged
Normal exit:   7 iterations. Status=0, F=    .1434121D-02
Maximum of   25 iterations. Exit iterations with status=1
Maximum of   25 iterations. Exit iterations with status=1
Iterative procedure has converged
Normal exit:  13 iterations. Status=0, F=    .8732352D-01
Iterative procedure has converged
Normal exit:   7 iterations. Status=0, F=    .8737990D-01
Iterative procedure has converged
Normal exit:   7 iterations. Status=0, F=    .8738937D-01
Iterative procedure has converged
Normal exit:   4 iterations. Status=0, F=    .8739076D-01
-----------------------------------------------------------------------------
Random Parameters Logit (BLP)  Model
-----------------------------------------------------------------------------

--------+--------------------------------------------------------------------
        |                  Standard            Prob.      95% Confidence
  SHARES|  Coefficient       Error       z    |z|>Z*         Interval
--------+--------------------------------------------------------------------
        |Variables with Nonrandom Parameters in Utility Function.............
Constant|   -1.18826       265.6695      .00  .9964  -521.89096  519.51444
      X1|     .92254***      .31271     2.95  .0032      .30965   1.53543
        |Variables with Random Parameters in Utility Function................
      X2|     .92450        1.01950      .91  .3645    -1.07369   2.92269
      X3|    -.14799        1.21830     -.12  .9033    -2.53581   2.23983
        |Standard Deviations of Random Parameters............................
     sX2|     .48469         .93317      .52  .6035    -1.34430   2.31368
     sX3|     .01339         .06622      .20  .8397     -.11640    .14318
--------+--------------------------------------------------------------------
***, **, * ==>  Significance at 1%, 5%, 10% level.
Model was estimated on Oct 11, 2016 at 07:41:08 PM
-----------------------------------------------------------------------------

这个软件5000伪币不知有没有人要。

贵到也不贵

不过大家用limdep软件的少
国内应该很少有学校用

学习了，谢谢

有没有安装包啊？