Im, Pesaran and Shin Panel Unit Root Test
res = IPS1(Y,t_order,pmax,signif)
* Y = matrix (T,N) of observations
The data matrix may be unbalanced.
Missing Values must be specified as %nan
* t_order = -1 : no individual effect
0 : individual effects (Default)
1 : individual effects and time trends
* pmax = Maximum of the lag order authorized
* signif = significance level for the individual ADF tests (0.05, 0.01 or 0.1)
* res = a results tlist with:
- res('meth') = 'IPS'
- res('y') = (T x k) matrix of data
- res('t_order') = the trend order (-1, 0 or 1)
- res('tbar') = Mean of Individual Augmented Dickey Fuller statistics
- res('Wbar') = Standardized IPS statistic based on E[ti(pi,0)/bi=0] and V[ti(pi,0)/bi=0]
- res('Wbar_pvalue') = Pvalue of Wbar
- res('Zbar') = Standardized IPS statistic based on the moments of the DF distribution
- res('Zbar_pvalue') = Pvalue of Zbar
- res('critical') = Critical Values of the Normal distribution at 1%, 5% and 10%
- res('tbar_DF') = Mean of Individual Dickey Fuller statistics
- res('Zbar_DF') = Standardized IPS statistic (assumption of no autocorrelation of residuals)
- res('Zbar_DF_pvalue') = Pvalue of Zbar_DF
- res('pi') = Individual lag order in individual ADF models
- res('pmax') = Maximum of lag order in individual ADF models
- res('Ti') = Adjusted Individual Size
load(GROCERDIR+'/data/gdpan_oecd.dat'); Y=explone(dblist(GROCERDIR+'/data/gdpan_oecd.dat')); // extract matrix of data present in database res=IPS1(log(Y),0,5,0.05) // Example provides the IPS test for all Y columns, with a constant but not trend (arg # 2 = 0), the maximum number of lags set to 5 (arg # 3 = 5). | ![]() | ![]() |