search for: sigma_theta

...t_5). Consider these 'longitudinal' data. So, I use reshape to put into the long format in order to regress the observations onto time, a time-varying covariate. I do the regression for each individual separately: Y_{t} = \mu + \beta(time) + \epsilon_{t} To get the statistic of interest (sigma_theta), I do the following with the original data: theta <- numeric(10) sigma <- numeric(10) for(i in 1:10){ tmp <- subset(long, ID==i) tmp.lm <- lm(obs ~ time, tmp) theta[i] <- coef(tmp.lm)[2] sigma[i] <- sum((tmp.lm$residuals)^2)/ (dim(tmp)[1] -2) } sig...

CRAN (and crantastic) updates this week New packages ------------ * disclapmix (0.1) Maintainer: Mikkel Meyer Andersen Author(s): Mikkel Meyer Andersen and Poul Svante Eriksen License: GPL-2 http://crantastic.org/packages/disclapmix disclapmix makes inference in a mixture of Discrete Laplace distributions using the EM algorithm. * EstSimPDMP (1.1) Maintainer: Unknown Author(s):