Hello R users, before asking my question I'd like to stress that I'm an (absolute) beginner in using R, but enthused about the incredible possibilities of it. So I hope my questions are not too stupid. Here's my problem: I have a dataset with skewed distributions. In order to obtain approx multivariate normality by a Box-Cox-Transformation, I used the "Rfwdmv Package" to estimate a common value of lambda for all the variables. Therefore I entered the following R commands: data(my.dat) l.mle <- fwdtr(my.dat, lambda = 1/3, one.lambda = TRUE) l.profile.mle <- profile.fwdtr(l.mle) fwdtrProfilePlot(l.profile.mle) fwdtrMlePlot(l.mle). Everything worked well so far and I obtained the plots in question. The problem is: In these plots I just can see that lambda is roughly around 1.4, but it could be 1.35 or 1.45 or somewhat in between. The same problem occurs with the log-likelihoods and the confidence intervals. Now my question is: Is there a possibility obtaining the EXACT log-likelihoods, confidence intervals and, first of all, lambda values ? Or more general: How can I obtain the x (or y) value for a specific (e.g. maximum) y (or x) of a plot in R ? I would be very grateful if one could give me a lead. Thank you very much in advance. With best regards, Juergen
Hello R users, before asking my question I'd like to stress that I'm an (absolute) beginner in using R, but enthused about the incredible possibilities of it. So I hope my questions are not too stupid. Here's my problem: I have a dataset with skewed distributions. In order to obtain approx multivariate normality by a Box-Cox-Transformation, I used the "Rfwdmv Package" to estimate a common value of lambda for all the variables. Therefore I entered the following R commands: data(my.dat) l.mle <- fwdtr(my.dat, lambda = 1/3, one.lambda = TRUE) l.profile.mle <- profile.fwdtr(l.mle) fwdtrProfilePlot(l.profile.mle) fwdtrMlePlot(l.mle). Everything worked well so far and I obtained the plots in question. The problem is: In these plots I just can see that lambda is roughly around 1.4, but it could be 1.35 or 1.45 or somewhat in between. The same problem occurs with the log-likelihoods and the confidence intervals. Now my question is: Is there a possibility obtaining the EXACT log-likelihoods, confidence intervals and, first of all, lambda values ? Or more general: How can I obtain the x (or y) value for a specific (e.g. maximum) y (or x) of a plot in R ? I would be very grateful if one could give me a lead. Thank you very much in advance. With best regards, Juergen