arlionn 发表于 2013-5-25 17:39
你的数据是 Panel Data ,而非 Time Series,所以不应用这个命令。而应采用 xtdes 命令描述样本的分布。
在 ...
多谢老师回复。
是这样,数据已经整理为panel data了。xtdes 描述对1071个T数据查起来还是不方便。
不过还是没法用 madfuller, pescadf ,ipshin 等命令查unit root, 或者说结果很奇怪,log 文件贴在下面,还请连老师给看看。感兴趣的变量是 Rate 和OCR。。
====================================
use try01.dta,clear
. xtset NameID Date
panel variable: NameID (strongly balanced)
time variable: Date, 24feb1992 to 29jan1995
delta: 1 day
. madfuller Rate, lags(5)
Multivariate Augmented Dickey-Fuller test for Rate
with 1071 observations on 1 cross-sectional units
-----------------------------------------------------------
Obs Lags MADF Approx 5% CV
-----------------------------------------------------------
1066 5 5.936 11.783
-----------------------------------------------------------
H0: all 1 timeseries in the panel are I(1) processes
. do "C:\Users\DONGXI~1\AppData\Local\Temp\STD00000000.tmp"
. pescadf Rate,lags(2)
Pesaran's CADF test for Rate
Cross-sectional average in first period extracted and extreme t-values truncated
Deterministics chosen: constant
t-bar test, N,T = (1,1071) Obs = 1068
Augmented by 2 lags (average)
t-bar cv10 cv5 cv1 Z[t-bar] P-value
2.610 . . . . .
.
end of do-file
. do "C:\Users\DONGXI~1\AppData\Local\Temp\STD00000000.tmp"
. pescadf Rate,lags(7)
Pesaran's CADF test for Rate
Cross-sectional average in first period extracted and extreme t-values truncated
Deterministics chosen: constant
t-bar test, N,T = (1,1071) Obs = 1063
Augmented by 7 lags (average)
t-bar cv10 cv5 cv1 Z[t-bar] P-value
2.610 . . . . .
.
end of do-file
. do "C:\Users\DONGXI~1\AppData\Local\Temp\STD00000000.tmp"
. ipshin Rate,lag(7)
Im-Pesaran-Shin test for cross-sectionally demeaned Rate
Deterministics chosen: constant
t-bar test, N,T = (1,1071) Obs = 1063
Augmented by 7 lags (average)
t-bar cv10 cv5 cv1 W[t-bar] P-value
. . . . . .
.
end of do-file