Sorry. Let me modify the question: Does there exist any unit root test function (with trend or intercept) where the p-value can be extracted? The function adf.test in tseries package does return the p-value, but there's no choice of trend or intercept. Thanks.
2015-03-13 10:49 GMT+08:00 jpm miao <miao...@gmail.com>: > Thank you very much. > > Could we extract the p-value in the output of the ur.df function? Does > there exist any unit root test function where the p-value can be extracted? > Thanks! > > An example for ur.df function: > > data(Raotbl3) > attach(Raotbl3) > lc.df <- ur.df(y=lc, lags=3, type='trend') > summary(lc.df) > > 2015-03-13 1:23 GMT+08:00 David Winsemius <dwinsem...@comcast.net>: > >> >> On Mar 12, 2015, at 1:04 AM, jpm miao wrote: >> >> > Hi, >> > >> > I run a statistical test function in the package "fUnitRoots" that >> > returns a S4 object but I am wondering how to extract the p-value, one >> of >> > the output elements. >> >> > The document of the function "urdfTest": >> > ..... >> > All tests return an object of class "fHTEST" with the following slots: >> > >> > @call >> > ..... >> > @test >> > a list object which holds the output of the underlying test function. >> > @title >> > ..... >> > The entries of the @test slot include the following components: >> > >> > $statistic >> > ...... >> > $p.value >> > the p-value of the test. >> > ..... >> > (end) >> > >> > I store the result of the test test in an element of a list >> > (adf1["r3m"][[1]]), and I want to extract the p-value. I was expecting >> the >> > p-value via adf1["r3m"][[1]]@test$p.value but it gives only an error >> > message. Could someone tell me how to extract the p-value? Thanks!!! >> >> If you look at the code by typing: fUnitRoots::urdfTest, you should >> quickly see why you are seeing text output: >> >> That function is doing the equivalent of a console screenscrape: >> >> ... >> output = capture.output(summary(urca))[-(1:4)] >> ... >> # and then assigns a trimmed version of that result to the `test`- slot. >> >> >> So the documentation is misleading in suggesting that a list object is >> being returned in the `test` slot. It's just a character vestor. >> >> >> -- >> > >> >> adf1["r3m"][[1]]<-urdfTest(dat[,i], lags = 1, type = "ct") >> >> adf1["r3m"][[1]] >> > [1] " " >> > [2] " Test regression trend " >> > [3] " " >> > [4] " Call:" >> > [5] " lm(formula = z.diff ~ z.lag.1 + 1 + tt + z.diff.lag)" >> > [6] " " >> > [7] " Residuals:" >> > [8] " Min 1Q Median 3Q Max " >> > [9] " -3.0785 -0.0485 0.0072 0.0627 3.5672 " >> > [10] " " >> > [11] " Coefficients:" >> > [12] " Estimate Std. Error t value Pr(>|t|) " >> > [13] " (Intercept) -1.733e-02 9.529e-03 -1.818 0.06910 . " >> > [14] " z.lag.1 -7.060e-03 2.343e-03 -3.013 0.00261 ** " >> > [15] " tt 5.299e-06 4.927e-06 1.076 0.28221 " >> > [16] " z.diff.lag -1.035e-01 1.859e-02 -5.569 2.81e-08 ***" >> > [17] " ---" >> > [18] " Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1" >> > [19] " " >> > [20] " Residual standard error: 0.2107 on 2864 degrees of freedom" >> > [21] " Multiple R-squared: 0.01461,\tAdjusted R-squared: 0.01358 " >> > [22] " F-statistic: 14.15 on 3 and 2864 DF, p-value: 3.716e-09" >> > [23] " " >> > [24] " " >> > [25] " Value of test-statistic is: -3.0134 3.0697 4.5828 " >> > [26] " " >> > [27] " Critical values for test statistics: " >> > [28] " 1pct 5pct 10pct" >> > [29] " tau3 -3.96 -3.41 -3.12" >> > [30] " phi2 6.09 4.68 4.03" >> > [31] " phi3 8.27 6.25 5.34" >> > >> >> That is being displayed as a multi-element text object. You instead need >> to show the code that actually created that object. >> >> >> adf1["r3m"][[1]]@test >> > Error: trying to get slot "test" from an object of a basic class >> > ("character") with no slots >> >> adf1["r3m"][[1]]@test$p-value >> > Error: trying to get slot "test" from an object of a basic class >> > ("character") with no slots >> > >> > [[alternative HTML version deleted]] >> >> Please learn to post in palin text. >> >> -- >> >> David Winsemius >> Alameda, CA, USA >> >> > [[alternative HTML version deleted]] ______________________________________________ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.