similar to: Inverse X'WX matrix from weighted linear regression

Hi list, can I extract the cov.unscaled ("the unscaled covariance matrix") from a gls fit (package nlme), like with summary.lm? Background: In a fixed effect meta analysis regression the standard errors of the coefficients can be computed as sqrt(diag(cov.unscaled)) where cov.unscaled is (X'WX). I try do do this with a gls-fit. Thanks, Sven

I am using the ginv function from MASS and have run across this problem that I do not understand. If I define the matrix A as below, its g-inverse does not satisfy the Moore-Penrose condition A %*% ginv(A) %*% A = A. The matrix A is X'WX in a quadratic regression using some very large dollar values. The much simpler matrix B does satisfy the MP condition. Am I doing something wrong? Is

Hello, I am working on a variable selection problem and I wonder whether there is some function or package in R works similar to the 'PROC REG' in SAS? Thank you. Some facts about 'PROC REG': PROC REG in SAS first composes a crossproducts matrix. The matrix can be calculated from input data, reformed from an input correlation matrix, or read in from an SSCP data set. For each

Hi dear R-users I try to reproduce the steps included in a LDA. Concerning the eigenvectors there is a difference to SPSS. In my textbook (Bortz) it says, that the matrix with the eigenvectors V usually are not normalized to the length of 1, but in the way that the following holds (SPSS does the same thing): t(Vstar)%*%Derror%*%Vstar = I where Vstar are the normalized eigenvectors. Derror

Hi! I have used the Matrix package (Version: 0.995-10) successfully to obtain the OLS solution for a problem where the design matrix X is 44000x6000. X is very sparse (about 80000 non-zeros elements). Now I want to do WLS: (X'WX)^-1X'Wy I tried W=Diagonal(length(w),w) and wX=solve(X,W) but after various minutes R gives a not enough memory error (Im using a 64bit machine with 16Gigs

Dear R-help, I have a four-part question about regression, matrices, and sandwich package. 1) In the sandwich package, I would like to better understand the meat() function. >From the bread() documentation, for a simple OLS regression, bread() returns (1/n * X'X)^(-1) That is, for a simple regression (per the documentation on bread()): MyLM <- lm(y ~ x) bread(MyLM)

Hi there Sorry for what may be a naive or dumb question. I have the following data: > x <- c(1,2,3,4) # predictor vector > y <- c(2,4,6,8) # response vector. Notice that it is an exact, perfect straight line through the origin and slope equal to 2 > error <- c(0.3,0.3,0.3,0.3) # I have (equal) ``errors'', for instance, in the measured responses Of course the

Hi all, how can I change the background color in lattice strips according to a factor level, eg: library(lattice) x <- rnorm(100) y <- sqrt(x) f <- gl(2, 50, c("A", "B")) xyplot(y ~ x | f) I like to change the background color of the strips according to the levels in f and tried several things like this with no success: xyplot(y ~ x | f,

Hi everyone, I admit I am a bit of an R novice, and I was hoping someone could help me with this error message: Warning message: In lm.fit(x, y, offset = offset, singular.ok = singular.ok, ...) : extra arguments weigths are just disregarded. My equation is: lm( Y ~ X1 + X2 + X3, weigths = seq(0.1, 1, by = 0.1)) -- View this message in context:

Dear list members, I have problems to interpret the coefficients from a lm model involving the interaction of a numeric and factor variable compared to separate lm models for each level of the factor variable. ## data: y1 <- rnorm(20) + 6.8 y2 <- rnorm(20) + (1:20*1.7 + 1) y3 <- rnorm(20) + (1:20*6.7 + 3.7) y <- c(y1,y2,y3) x <- rep(1:20,3) f <- gl(3,20,

I would like to apply network measures (such as betweenness centrality, upper boundedness, etc.) to a weighted graph with non-integer weights, defined by a euclidean distance matrix. The package sna provides the measures that I want to use, but seems only to operate on binary graphs. I have read work by Mark Newman (http://aps.arxiv.org/abs/cond-mat/0407503/), who suggests that a weighted graph

Hi all, how do I change the magnification of xaxis annotation? I need a smaller axis text size in some of my plots. I tried cex.axix=0.5 in plot(), but this doesn't work. Thanks, Sven -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or

Hi all, I don't understand the behavior of is.integer(): > x <- integer() > is.integer(x) [1] TRUE > x <- 10 > is.integer(x) [1] FALSE > x <- 1:10 > is.integer(x) [1] TRUE Why is.interger() returns FALSE if x has only one element? And how can someone check if x is an integer but contains only one value? (R 1.5.0 on Linux i386) Thanks, Sven

Hi, I am working on a project about hospital efficiency. Due to the high multicolinearlity of the data, I want to fit the model using ridge regression. However, I believe that the data from large hospital(indicated by the number of patients they treat a year) is more accurate than from small hosptials, and I want to put more weight on them. How do I do this with lm.ridge? I know I just need

I understand that the hat matrix is a function of the predictor variable alone. So, in the following example why do the values on the diagonal of the hat matrix change when I go from an unweighted fit to a weighted fit? Is the function hatvalues giving me something other than what I think it is? library(ISwR) data(thuesen) attach(thuesen) fit <- lm(short.velocity ~ blood.glucose)

Hi all, there is a vector v with several NAs. I want to create a new vector n of the same length as v and the same NAs as in v and tried this: n <- vector(length=length(v), mode="numeric") replace(n, which(is.na(v)), NA) but this does't work, all values in n are 0. What went wrong? Thanks, Sven -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-

Hi all, suppose I've got a vector y with some data (from a repeated measure design) observed given the conditions in f1 and f2. I've got a model with two unknown fix constants a and b which tries to predict y with respect to the values in f1 and f2. Here is an exsample # "data" y <- c(runif(10, -1,0), runif(10,0,1)) # f1 f1 <- rep(c(-1.4, 1.4), rep(10,2)) # f2 f2 <-

Hi there, if I copy an x11() graphics device to an eps-file (with dev.copy2eps()) the font in the legend is very large and doesn't fit to the legend box in the eps-file (same with a postscript file). I'm not sure if this is a R problem rather than a ghostscript one. But is there a way to solve this problem in R or depends this on my ghostscript installation? System: R Version 1.2.3 on

Hi, This is more of a general question than a pure R one, but I hope that is OK. I want to combine one-tailed independent p-values using the weighted version of fisher's inverse chi-square method. The unweighted version is pretty straightforward to implement. If x is a vector with p-values, then I guess that this will do for the unweighted version: statistic <- -2*sum(log(x)) comb.p <-

Hi there, when plotting a time series I got this error message: Error in xy.coords(x, y, xlabel, ylabel, log) : x and y lengths differ In addition: Warning messages: 1: longer object length is not a multiple of shorter object length in: vx.1[int.lr] + (-1) * vx.2[int.rl] vx.1[int.rl] and vx.2[int.rl] have the same length. Does anybody know what the warning message means? Thanks, Sven