tag:blogger.com,1999:blog-623696624102177636.post7686785748202053424..comments2017-01-03T05:28:57.067-08:00Comments on Anything but R-bitrary: Random regression coefficients using lme4Chris Nicholashttp://www.blogger.com/profile/09479130398118611479noreply@blogger.comBlogger9125tag:blogger.com,1999:blog-623696624102177636.post-65977523596269374732016-11-06T08:40:49.891-08:002016-11-06T08:40:49.891-08:00Fixed. Thanks!Fixed. Thanks!Ben Ogorekhttp://www.blogger.com/profile/16005153347460476695noreply@blogger.comtag:blogger.com,1999:blog-623696624102177636.post-51518152706905783432016-11-01T12:51:03.404-07:002016-11-01T12:51:03.404-07:00Nice post. A typo: "courser grain" shou...Nice post. A typo: "courser grain" should be "coarser grain"Michael Nealehttp://www.blogger.com/profile/04393720419103422688noreply@blogger.comtag:blogger.com,1999:blog-623696624102177636.post-18497232594402973162016-01-23T19:21:38.142-08:002016-01-23T19:21:38.142-08:00Hello! I am trying to use your code for separate r...Hello! I am trying to use your code for separate regressions but am having a little bit of trouble adjusting it as I would like to include a quadratic term into my lm and would like to estimate the coefficient for both the linear term and the quadratic term. Do you have any advice on how to expand it to include a quadratic term? Thanks!<br /><br />Ex) (animal id is my unit, x=fitness(hazard ratio), y=pathogen burden)<br />beta.hat <- list()<br />for(i in 1:nrow(data)){<br /> id.x.lm <- lm(hazardratio ~ burden+I(burden^2), data = subset(data, id.x == i) )<br /> beta.hat[i] <- coef(id.x.lm)[2]<br />}<br />beta.hat <- as.numeric(beta.hat)<br />Unknownhttp://www.blogger.com/profile/01179665048006688106noreply@blogger.comtag:blogger.com,1999:blog-623696624102177636.post-4719293339399985602015-01-02T11:13:50.164-08:002015-01-02T11:13:50.164-08:00I think the killer app that makes lmer models easy...I think the killer app that makes lmer models easy to interpret is the lsmeans package by Russell Lenth. I've found it really helpful; it filled some major gaps for me.daveCooperhttp://www.blogger.com/profile/04460434002817476914noreply@blogger.comtag:blogger.com,1999:blog-623696624102177636.post-65574923076964784462014-01-03T15:22:24.688-08:002014-01-03T15:22:24.688-08:00Thanks a lot for the great post! A question: as lm...Thanks a lot for the great post! A question: as lme4 does not support mcmcsamp() anymore, do you have any suggestion for doing MCMC sampling and estimating confidence intervals? Thanks!Danilohttp://www.blogger.com/profile/13334532726460175116noreply@blogger.comtag:blogger.com,1999:blog-623696624102177636.post-12394624099477630942013-12-14T09:23:44.514-08:002013-12-14T09:23:44.514-08:00...if I get one observation from a 'new' u......if I get one observation from a 'new' unit (not in dataset), how do I get predicted values for other observations from that unit, including prediction intervals?kma9http://www.blogger.com/profile/09107031225645680757noreply@blogger.comtag:blogger.com,1999:blog-623696624102177636.post-49545994908435247552012-12-21T14:38:45.482-08:002012-12-21T14:38:45.482-08:00Hi Daniel,
As far as mixed effects models, th...Hi Daniel,<br /><br /> As far as mixed effects models, there's a brief history section at http://en.wikipedia.org/wiki/Mixed_model. I was not myself familiar with the name Charles Roy Henderson. As a statistician, REML estimation comes to mind, and the Wikipedia page http://en.wikipedia.org/wiki/Reml lists the classic references I'm familiar with.<br /><br />Cheers,<br />Ben<br /><br />Ben Ogorekhttp://www.blogger.com/profile/16005153347460476695noreply@blogger.comtag:blogger.com,1999:blog-623696624102177636.post-7954177046235000162012-11-30T21:12:17.299-08:002012-11-30T21:12:17.299-08:00Two comments on the same day!
Thank you very much...Two comments on the same day!<br /><br />Thank you very much for this very nice post. The detailled examples help an awlful lot to grasp the concepts.<br /><br />---jjapUnknownhttp://www.blogger.com/profile/12109677816605190146noreply@blogger.comtag:blogger.com,1999:blog-623696624102177636.post-33767737778248069232012-11-30T09:37:55.848-08:002012-11-30T09:37:55.848-08:00Dear Ogorek!
Thank you for sharing this informati...Dear Ogorek!<br /><br />Thank you for sharing this information. By the way, what is the classical reference for these models? I mean what is the original paper, which is behind the estimation of lm4? <br /><br />DanieloTheMusicManhttp://www.blogger.com/profile/16699474896456088525noreply@blogger.com