[R] help generalizing the following non-symmetric "identity"-like matrix

Hi, I need help generalizing the following non-symmetric "identity"-like matrix In this case: p = 5 k=3 and I want this: I.kp <- cbind(diag(3), c(0, 0, 0), c(0, 0, 0)) How can I generalize this so that I don't have to add/subtract c(0, 0, ... , 0) when p and k changes? Hopefully that makes sen

Re: [R] MASS mve_fits mycov.rob

rseM_1.03 > [5] MASS_7.3-28 lattice_0.20-15 stringr_0.6.2 reshape2_1.2.2 > > loaded via a namespace (and not attached): > [1] magic_1.5-4 Matrix_1.0-12 mgcv_1.7-24 nlme_3.1-110 plyr_1.8 > [6] tcltk_3.0.1 tools_3.0.1 > > A.K. > > > > > > > - Ori

[R] MASS mve_fits mycov.rob

Hello, In R 3.0.1, I get the following warning that I do not get in R 2.15.3: data(mtcars) mycov.rob(mtcars[,1:3], method="mcd") Error in .C("mve_fitlots", as.double(x), as.integer(n), as.integer(p), : "mve_fitlots" not available for .C() for package "MASS" It seems like there was a change in

[R] Simulated mixed distribution multivariate data

Hi, I want to simulate multivariate data with > 1 distribution type. For example, I would like one normal variate and one poisson variate with a specified correlation structure. Is there a package that has this implemented? Thanks. Chuck [[alternative HTML version deleted]]

Re: [R] Same plot size with mfrow()

13 at 6:33 PM, Bert Gunter wrote: > It is hard to know what to recommend, as you don't give us any context. > > If the plots are interrelated in certain ways, lattice and ggplot both > provide conditioning plots.Check vignettes and documentation for these > packages. > >

[R] Same plot size with mfrow()

Hello, I want to plot similar to the example below. I want to remove the X and Y axis labels to avoid redundancy, but then want to make the spaces between plots smaller. When I do this (see below), I end up with plots of different sizes (I want the plots to be exactly the same dimensions). Coul