高手们好，菜鸟刚处理数据就遇到问题了。
一时间序列，平稳性检验ADF时，么有AIC值该怎么确定呢。。
我分别用的滞后一期、滞后二期、滞后三期、滞后四期看滞后期的系数的t值或P值判断呢？似乎应该是选用滞后三期。
dfuller r, trend regress lag(1)
Augmented Dickey-Fuller test for unit root Number of obs = 23795
---------- Interpolated Dickey-Fuller ---------
Test 1% Critical 5% Critical 10% Critical
Statistic Value Value Value
------------------------------------------------------------------------------
Z(t) -109.900 -3.960 -3.410 -3.120
------------------------------------------------------------------------------
MacKinnon approximate p-value for Z(t) = 0.0000
------------------------------------------------------------------------------
D.r | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
r |
L1. | -1.002621 .009123 -109.90 0.000 -1.020503 -.9847398
LD. | .0109783 .0064728
1.70 0.090 -.0017089 .0236655
_trend | 6.36e-10 1.61e-09 0.40 0.693 -2.52e-09 3.79e-09
_cons | -.0000184 .0000221 -0.83 0.404 -.0000618 .0000249
------------------------------------------------------------------------------
. dfuller r, trend regress lag(2)
Augmented Dickey-Fuller test for unit root Number of obs = 23794
---------- Interpolated Dickey-Fuller ---------
Test 1% Critical 5% Critical 10% Critical
Statistic Value Value Value
------------------------------------------------------------------------------
Z(t) -88.937 -3.960 -3.410 -3.120
------------------------------------------------------------------------------
MacKinnon approximate p-value for Z(t) = 0.0000
------------------------------------------------------------------------------
D.r | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
r |
L1. | -.9960569 .0111996 -88.94 0.000 -1.018009 -.974105
LD. | .00449 .0091211 0.49 0.623 -.013388 .0223681
L2D. | -.0077242 .0064719
-1.19 0.233 -.0204096 .0049611
_trend | 5.72e-10 1.61e-09 0.36 0.722 -2.58e-09 3.73e-09
_cons | -.0000174 .0000221 -0.79 0.432 -.0000607 .000026
------------------------------------------------------------------------------
dfuller r, trend regress lag(3)
Augmented Dickey-Fuller test for unit root Number of obs = 23793
---------- Interpolated Dickey-Fuller ---------
Test 1% Critical 5% Critical 10% Critical
Statistic Value Value Value
------------------------------------------------------------------------------
Z(t) -76.893 -3.960 -3.410 -3.120
------------------------------------------------------------------------------
MacKinnon approximate p-value for Z(t) = 0.0000
------------------------------------------------------------------------------
D.r | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
r |
L1. | -.9940608 .0129278 -76.89 0.000 -1.0194 -.9687215
LD. | .0027262 .0111993 0.24 0.808 -.0192252 .0246776
L2D. | -.009473 .009121 -1.04 0.299 -.0273507 .0084047
L3D. | -.0011161 .006472 -0.17 0.863 -.0138016 .0115693
_trend | 6.03e-10 1.61e-09 0.37 0.708 -2.55e-09 3.76e-09
_cons | -.0000178 .0000221 -0.81 0.420 -.0000612 .0000255
------------------------------------------------------------------------------
dfuller r, trend regress lag(4)
Augmented Dickey-Fuller test for unit root Number of obs = 23792
---------- Interpolated Dickey-Fuller ---------
Test 1% Critical 5% Critical 10% Critical
Statistic Value Value Value
------------------------------------------------------------------------------
Z(t) -67.724 -3.960 -3.410 -3.120
------------------------------------------------------------------------------
MacKinnon approximate p-value for Z(t) = 0.0000
------------------------------------------------------------------------------
D.r | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
r |
L1. | -.9782247 .0144443 -67.72 0.000 -1.006536 -.949913
LD. | -.0131251 .0129268 -1.02 0.310 -.0384624 .0122122
L2D. | -.0254523 .0111984
-2.27 0.023 -.0474019 -.0035027
L3D. | -.0169192 .0091205
-1.86 0.064 -.0347959 .0009574
L4D. | -.0159508 .0064714
-2.46 0.014 -.0286352 -.0032664
_trend | 5.90e-10 1.61e-09 0.37 0.714 -2.57e-09 3.75e-09
_cons | -.0000175 .0000221 -0.79 0.429 -.0000609 .0000259
------------------------------------------------------------------------------
我是用他们的滞后期的系数的显著性选择 滞后2期。。因为四个模型的r都是平稳的，但是标红的系数却不够显著。。所以就干脆用lag(2)。。。我这样理解对么？或者该怎么用ADF检验的时候给出AIC值呢？
之前有人说dfgls命令，
https://bbs.pinggu.org/thread-660053-1-1.html
但是这个适合时序经过GLS处理过的。。我的也没处理过吧。。
各种不懂，拜谢各位了。
扫码加好友，拉您进群
全部版块
我的主页
收藏
