Library: | times |
See also: | rvlm lo robwhittle roblm lobrob gph sclmtest neweywest |
Macro: | kpss | |
Description: |
Calculation of the KPSS statistics for I(0) against long-memory
alternatives. We consider the two tests, denoted by KPSS_mu and KPSS_t,
and respectively based on a regression on a constant mu, and
on a constant and a time trend t. As in the KPSS paper, we
consider for the autocorrelation consistent variance
estimator the truncation lags denoted by L0, L4, and L12.
The quantlet returns the value of the truncation lag, the
type of the test, i.e., const or trend, the estimated statistic,
and the critical value for a 95 percent confidence interval for I(0).
If the value of the test exceeds the critical value, a star symbol
* is displayed after the test statistic.
|
Usage: | k = kpss(y) | |
Input: | ||
y | vector | |
Output: | ||
k | vector |
;Non-stationarity test on the series dmus58.dat. Since the evaluated ;statistics are over the critical value for I(0), for all the ;truncation orders, the star symbol is displayed after all the ;computed statistics. library("times") x = read("dmus58.dat") ; 25477 obs: log FX rates x = x[1:1000] k = kpss(x) ; k
Contents of k [1,] " Order Test Statistic Crit. Value " [2,] "_______________________________________ " [3,] "" [4,] " L0 = 0 const 77.7016 * 0.4630" [5,] " L4 = 7 const 9.7786 * 0.4630" [6,] " L12= 21 const 3.5996 * 0.4630" [7,] " L0 = 0 trend 11.1167 * 0.1460" [8,] " L4 = 7 trend 1.4222 * 0.1460" [9,] " L12= 21 trend 0.5412 * 0.1460"
Library: | times |
See also: | rvlm lo robwhittle roblm lobrob gph sclmtest neweywest |