[求助]在stata里如何简单快捷的用一个var过程生成新数据

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

如题，现在有y，z两个变量，我想生成x变量，x变量用一个x,y,z的var过程来生成，像这样，x_t=a_0+a_11*x_t-1+a_12*y_t-1+a_13*z_t-1+a_21*x_t-2+a_22*y_t-2+a_23*z_t-2+epsilon(t)，具体的操作命令是什么啊？
才开始学stata不久，他的命令很不熟，用惯了像matlab之类的软件，语法习惯也不习惯，郁闷，谁能指导下如何更有效快速的学习stata

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏0 回帖

关键词：Stata tata VaR MATLAB matla matlab 如何 软件

dd0627 发表于 2010-5-27 20:00 现在有y，z两个变量，我想生成x变量，x变量用一个x,y,z的var过程来生成，像这样，
x_t=a_0+a_11*x_t-1+a_12*y_t-1+a_13*z_t-1+a_21*x_t-2+a_22*y_t-2+a_23*z_t-2+epsilon(t)

*是不是要以下结果？（设t是时间标识，x的初二值都是0）

tsset t
g x=.
replace x=0 in 1/2
replace x=a0+a11*L.x+a12*L.y+a13*L.z+a21*L2.x+a22*L2.y+a23*L2.z+uniform() in 3/l

1. tsset:
       tsset x t, yearly (or monthly et al.)                    * suppose x is the panel var

2. gen a NEW var (you may choose not to use "replace" per sungmoo's suggestion)
        gen x_new=a0+a11*L.x+a12*L.y+a13*L.z+a21*L2.x+a22*L2.y+a23*L2.z+uniform()

johnayl 发表于 2010-5-28 05:11
1. tsset:
       tsset x t, yearly (or monthly et al.)                    * suppose x is the panel var

2. gen a NEW var (you may choose not to use "replace" per sungmoo's suggestion)
        gen x_new=a0+a11*L.x+a12*L.y+a13*L.z+a21*L2.x+a22*L2.y+a23*L2.z+uniform()

dd0627 发表于 2010-5-27 20:00 如题，现在有y，z两个变量，我想生成x变量，x变量用一个x,y,z的var过程来生成

NOTE: in this question, x has not been generated.

Thanks for Sungmoo and johnayl, I have found the mistake i made. A display of my mistake as follows
set sample 100
t=_n
tsset t
gen epsilon=invnormal(uniform())
gen x=0
replace x=a0+a11*L.x+a12*L.y+a13*L.z+epsilon
The commands above have a mistake, since x[1]=f(x[0]), however x[0] doesn't even exist, thus stata will replace x series with wrong elements, the right commands are below
replace x=a0+a11*L.x+a12*L.y+a13*L.z+epsilon if t>1