About Me
My name is Christopher David Desjardins. I am a Ph.D. student studying quantitative methods in education in the department of educational psychology at the University of Minnesota. I am also pursing a M.S. in statistics from the University of Minnesota. I am a Minnesota Interdisciplinary Training in Education Research fellow. I am interested in Bayesian statistics, personal probabilities and causal inferences, longitudinal multilevel models, latent variable modeling (specifically structural equation modeling and factor analysis), free software, Debian GNU/Linux, R , and psychology.
I have installed Winbugs using your directions. It says that the licence expired in 2004.
Did you patch WinBUGS with this?
http://www.mrc-bsu.cam.ac.uk/bugs/winbugs/WinBUGS14_immortality_key.txt
Dear Christopher Desjardins,
I could not locate your email address so I hope you don’t mind the comment.
I have created an evaluation news site that links together feeds from evaluation blogs across the web.
As a starting point, I aggregated RSS/Atom feeds from the American Evaluation Association: Evaluation Blogs list http://www.eval.org/Resources/Blogs.asp Your blog was on this list, so it was included.
Eval Central has just been set up and is not published to search engines. Additionally, there has been no attempt to advertise it’s existence. Before doing so, I am contacting administrators for all the blogs used during the development phase.
The site has four goals:
Create a destination news source for readers interested in evaluation.
Increase traffic to all participating blogs, with little to no attached burden.
Provide an opportunity for low post frequency bloggers to reach a larger audience.
Provide an opportunity for evaluators without a blog to easily post to a site with an established audience.
This is an academic venture on my part. I believe that if we can increase the frequency and diversity of content easily available to evaluators, and make it easier to post evaluation content for an established audience, we can increase the amount of high quality evaluation content available online.
Here is the site: http://evalcentral.wordpress.com/
You’re welcome to add me. But it should be noted that I am not blogging very often and most of the posts that I do are about statistics.
Can you point me to a good example (a paper or R code or both) of MCMCglmm used for modeling effects of hierarchical predictor variables on multinomial response variables?
thanks in advance,
pasky (pls. send response to my work e-mail)
Hi,
Have you tried the CourseNotes vignette that comes with MCMCglmm? They are really good and worth examining.
Christopher David Desjardins, i’m a student who study technology management, i want to measure the technology acceptance which my company’s develop but i have a problem with the number of sample, i only have 43 sample so i have been suggested to use bayesian sem, but i don’t understand much about statistic. i’m a newbie in bayesian sem, i try to write a code base on yours but i have a problem with my winbugs syntax…could you help me? which one that i used fix? i will very grateful for your help…
model{
for(i in 1:N){
#model persamaan pengukuran
for(j in 1:P){
y[i,j]~dnorm(mu[i,j],psi [j]) I(thd [j,z[i,j]],thd[j,z[i,j]+1])
ephat[i,j]<-y[i,j] -mu[i,j]
}
#faktor Budaya Organisasi
mu[i,1]<-xi[i,1]
mu[i,2]<-lam[1]*xi[i,1]
mu[i,3]<-lam[2]*xi[i,1]
#faktor Kemampuan Pengguna
mu[i,4]<-xi[i,2]
mu[i,5]<-lam[3]*xi[i,2]
mu[i,6]<-lam[4]*xi[i,2]
#faktor Mekanisme Dukungan
mu[i,7]<-xi[i,3]
mu[i,8]<-lam[5]*xi[i,3]
mu[i,9]<-lam[6]*xi[i,3]
#faktor Desain Antarmuka
mu[i,10]<-xi[i,4]
mu[i,11]<-lam[7]*xi[i,4]
mu[i,12]<-lam[8]*xi[i,4]
#faktor Persepsi Kualitas
mu[i,13]<-xi[i,5]
mu[i,14]<-lam[9]*xi[i,5]
mu[i,15]<-lam[10]*xi[i,5]
#faktor Persepsi Kemudahan Kegunaan
mu[i,16]<-eta[i,1]
mu[i,17]<-lam[11]*eta[i,1]
mu[i,18]<-lam[12]*eta[i,1]
#faktor Persepsi Kegunaan
mu[i,19]<-eta[i,2]
mu[i,20]<-lam[13]*eta[i,2]
mu[i,21]<-lam[14]*eta[i,2]
mu[i,22]<-lam[15]*eta[i,2]
#faktor Sikap ke arah Penggunaan
mu[i,23]<-eta[i,3]
mu[i,24]<-lam[16]*eta[i,3]
mu[i,25]<-lam[17]*eta[i,3]
#faktor Persepsi Niat untuk Menggunakan
mu[i,26]<-eta[i,4]
mu[i,27]<-lam[18]*eta[i,4]
mu[i,28]<-lam[19]*eta[i,4]
#faktor Adopsi E-government
mu[i,29]<-eta[i,5]
mu[i,30]<-lam[20]*eta[i,5]
#model persamaan struktural
xi[i,1:5] ~dmnorm(u[1:5],phi[1:5,1:5])
eta[i,1]~dnorm(nu[i,1],pskp)
nu[i,1]<-gam[1]*xi[i,2]+gam[2]*xi[i,3]+gam[3]*xi[i,4]
dthat[i,1]<-eta[i,1]-nu[i,1]
eta[i,2]~dnorm(nu[i,2],pspk)
nu[i,2]<-gam[4]*xi[i,1]+beta[1]*eta[i,1]
dthat[i,2]<-eta[i,2]-nu[i,2]
eta[i,3]~dnorm(nu[i,3],pssp)
nu[i,3]<-beta[2]*eta[i,2]+beta[3]*eta[i,3]
dthat[i,3]<-eta[i,3]-nu[i,3]
eta[i,4]~dnorm(nu[i,4],psnm)
nu[i,4]<-beta[4]*eta[i,1]+beta[5]*eta[i,2]+gam[5]*xi[i,5]
dthat[i,4]<-eta[i,4]-nu[i,4]
eta[i,5]~dnorm(nu[i,5],psae)
nu[i,5]<-beta[6]*eta[i,4]
dthat[i,5]<-eta[i,5]-nu[i,5]
}#akhir dari i
for (i in 1:5) {u[i]<-0.0}
#lamda
var.lam[1]<-8.0*psi[2] var.lam[2]<-8.0*psi[3]
var.lam[3]<-8.0*psi[5] var.lam[4]<-8.0*psi[6]
var.lam[5]<-8.0*psi[8] var.lam[6]<-8.0*psi[9]
var.lam[7]<-8.0*psi[11] var.lam[8]<-8.0*psi[12]
var.lam[9]<-8.0*psi[14] var.lam[10]<-8.0*psi[15]
var.lam[11]<-8.0*psi[17] var.lam[12]<-8.0*psi[18] var.lam[13]<-8.0*psi[20]
var.lam[14]<-8.0*psi[21] var.lam[15]<-8.0*psi[22]
var.lam[16]<-8.0*psi[24] var.lam[17]<-8.0*psi[25]
var.lam[18]<-8.0*psi[27] var.lam[19]<-8.0*psi[28]
var.lam[20]<-8.0*psi[30]
for (i in 1:20) {lam[i] ~dnorm(1,var.lam[i])}
for (j in 1:P) {
psi[j] ~dgamma(10,8)
sgl[j]<-1/psi[j]
}
#gamma
gam[1]~dnorm(0.4,var.pk)
gam[2]~dnorm(0.4,var.kp)
gam[3]~dnorm(0.4,var.kp)
gam[4]~dnorm(0.4,var.kp)
gam[5]~dnorm(0.4,var.nm)
var.pk <-8.0*pspk pspk~dgamma(10,8) sgpk<-1/pspk
var.kp <-8.0*pskp pskp~dgamma(10,8) sgkp<-1/pskp
var.sp <-8.0*pssp pssp~dgamma(10,8) sgsp<-1/pssp
var.nm <-8.0*psnm psnm~dgamma(10,8) sgnm<-1/psnm
var.ae <-8.0*psae psae~dgamma(10,8) sgae<-1/psae
#beta
beta[1] ~dnorm(0.5,var.pk)
beta[2] ~dnorm(0.5,var.sp)
beta[3] ~dnorm(0.5,var.sp)
beta[4] ~dnorm(0.5,var.nm)
beta[5] ~dnorm(0.5,var.nm)
beta[6] ~dnorm(0.5,var.ae)
phi[1:5,1:5] ~dwish(R[1:5,1:5],21)
phx[1:5,1:5]<-inverse(phi[1:5,1:5])
}
#end of model
DATA
list(N=43, P=30,
R=structure(
.Data=c(10,0,0,0,0,
0,10,0,0,0,
0,0,10,0,0,
0,0,0,10,0,
0,0,0,0,10
),
.Dim=c(5,5)),
thd=structure(
.Data=c(-250,-1.99072046156042,-1.08241139430414,0.983052916910141,250,-250,-200,-1.47752529199845,0.452147411138078,250,-250,-200,-1.08241139430414,0.730448177619092,250,-250,-1.99072046156042,-0.585607161227169,1.08241139430414,250,-250,-1.99072046156042,-0.265404753825216,1.08241139430414,250,-250,-200,-0.656304990872144,0.892559673266593,250,-250,-200,-1.08241139430414,0.517723552818072,250,-250,-1.99072046156042,-1.47752529199845,0.585607161227169,250,-250,-200,-1.99072046156042,0.265404753825216,250,-250,-200,-1.67966118528897,0.656304990872144,250,-250,-1.99072046156042,-1.67966118528897,0.517723552818072,250,-250,-1.67966118528897,-0.892559673266593,0.730448177619092,250,-250,-200,-1.08241139430414,0.80884440410662,250,-250,-200,-1.67966118528897,0.730448177619092,250,-250,-1.67966118528897,-1.19379507272265,0.80884440410662,250,-250,-200,-1.32236537894944,0.98305291691014,250,-250,-200,-0.656304990872144,1.08241139430414,250,-250,-1.99072046156042,-1.32236537894944,0.585607161227169,250,-250,-200,-1.19379507272265,0.656304990872144,250,-250,-1.99072046156042,-1.67966118528897,0.80884440410662,250,-250,-1.99072046156042,-1.19379507272265,0.656304990872144,250,-250,-1.67966118528897,-1.19379507272265,0.452147411138078,250,-250,-200,-1.47752529199845,0.656304990872144,250,-250,-1.99072046156042,-1.47752529199845,0.585607161227169,250,-250,-200,-1.19379507272265,0.730448177619092,250,-250,-1.67966118528897,-0.80884440410662,0.892559673266593,250,-250,-1.99072046156042,-1.67966118528897,0.585607161227169,250,-250,-200,-0.983052916910141,0.80884440410662,250,-250,-200,-1.67966118528897,0.585607161227169,250,-250,-1.99072046156042,-1.08241139430414,0.892559673266593,250),
.Dim=c(30,5)),
z=structure(
.Data=c(3,3,2,3,2,2,4,3,3,3,3,3,3,3,3,3,3,3,3,2,3,3,3,3,3,3,3,3,3,2,3,3,3,3,3,2,2,3,3,3,3,3,2,3,3,3,3,2,3,2,2,2,2,2,2,2,3,3,3,3,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,2,2,3,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,3,4,3,3,3,3,3,2,3,4,2,4,4,4,4,3,3,3,3,2,2,3,3,3,3,3,3,3,3,3,3,3,3,4,4,4,2,2,3,2,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,4,3,3,3,3,3,3,3,3,3,2,2,2,3,3,3,3,3,4,3,2,3,3,3,3,4,3,2,4,3,3,3,3,3,3,3,3,3,2,3,2,2,2,2,3,3,4,3,3,2,3,3,3,3,2,4,3,3,4,4,3,3,2,3,3,2,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,2,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,2,3,3,3,2,3,4,4,2,2,3,3,4,4,4,4,3,3,3,3,3,4,3,3,3,3,3,4,4,4,3,4,4,4,4,3,3,3,3,3,3,4,3,3,3,3,2,3,3,1,4,4,4,2,3,3,3,3,3,3,1,3,3,3,3,2,3,3,2,2,2,3,3,3,3,3,3,3,3,3,3,3,3,3,2,3,3,3,3,3,3,3,2,3,2,4,3,2,2,2,2,3,3,3,3,3,2,3,3,3,3,3,3,3,3,3,3,3,3,3,2,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,2,2,2,2,2,3,3,3,4,4,4,4,3,3,2,4,3,3,3,2,3,4,2,4,4,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,2,3,3,3,3,3,3,2,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,2,3,3,3,3,2,4,4,4,3,3,3,3,3,3,3,3,4,4,3,3,3,3,4,3,1,4,3,4,3,3,3,3,3,3,3,3,3,3,3,3,3,2,3,3,3,3,3,3,2,3,3,3,3,3,2,3,2,3,2,3,4,3,3,3,3,3,4,4,3,3,3,3,3,3,3,3,3,3,3,4,3,3,3,4,3,3,4,3,3,3,3,3,3,2,2,3,3,4,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,2,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,4,4,3,4,4,3,4,4,4,4,3,3,3,3,3,3,3,3,3,4,4,3,3,3,3,3,3,3,3,3,2,3,4,3,2,3,1,3,3,3,3,2,2,2,3,3,2,3,2,2,2,4,4,4,2,3,2,3,3,2,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,2,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,2,3,2,2,3,3,2,3,3,2,2,3,3,3,3,3,3,3,2,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,4,4,3,3,2,2,1,1,2,2,2,2,2,1,1,2,2,2,1,1,1,2,2,2,2,2,2,2,2,2,3,2,1,1,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,3,4,4,3,3,4,4,3,3,4,4,4,3,3,3,4,4,4,3,3,4,4,4,4,4,4,4,4,4,4,3,3,3,2,2,3,4,4,4,3,3,3,4,4,4,3,3,4,3,3,3,3,3,3,3,3,4,3,4,3,3,3,3,3,3,3,2,3,3,3,3,3,3,3,2,3,3,3,3,2,3,3,3,3,3,3,3,2,3,3,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,3,4,3,3,3,3,3,3,4,3,3,3,4,4,4,3,4,3,3,3,3,4,3,3,3,3,3,3,3,3,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,3,3,3,3,3,3,3,3,3,3,3,2,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,2,3,3,4,3,2,3,4,2,3,4,3,2,1,3,4,1,3,2,1,3,2,1,2,2,1,2,2,1,2,3,4,2,2,3,2,2,2,3,3,3,3,3,3,3,3,3,3,3,3,3,2,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,4,3,3,2,3,4,4,4,4,4,4,3,3,4,4,4,4,3,3,4,4,4,4,3,3,4,3,4,3
),
.Dim=c(43,30)))
)
INITS
list(
lam=c(0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0),
psi=c(1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0),
pspk=1.0, pskp=1.0, pssp=1.0, psnm=1.0, psae=1.0,
gam=c(0,0,0,0,0),
beta=c(0,0,0,0,0,0),
phi=structure(.Data=c(1,0,0,0,0,
0,1,0,0,0,
0,0,1,0,0,
0,0,0,1,0,
0,0,0,0,1
),.Dim=c(5,5)),
xi=structure(.Data=c(0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0),
.Dim=c(43,5)),
eta=structure(.Data=c(0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0),
.Dim=c(43,5))
)
Hello Christopher,
I managed to install Winbugs following you instruction,
but I get the annoying black box trap #101 error.
after typing
wine ~/.wine/drive_c/Program Files/WinBUGS14/WinBUGS14.exe
What I have missed? Thank you
Anna
This could be caused by a different version of Wine. What version of Wine are you using? Also, may I suggest you consider either OpenBUGS or JAGS? Both of which are freely available for Linux?
Dear Christopher Desjardins:
I come to you because in a material you posted to your blog (
Bayesian, MCMC, and Multilevel Modeling (a foray Into the Subjective)) at the end you say: “If you need help with syntax Please ask.”
I explain my situation:
I’m working on the evolution of the social structure of a particular insect. I want to evaluate whether the colonial size(independent variable, continous) of the species predicts the social structure(variable discrete with 4 states) in such insects. for this I fit a GLMM in the program MCMCglmm of R, since my response variable is discrete with 4 states. the problem lies is not good as well adjusted prior to model.
the model I’m trying to implement is the following:
model <- MCMCglmm (EstructuraSOC ~ Colonysize -1,
random = ~ animal
rcov = ~ us (trait): units,
family = "categorical"
Prior = list (R = list (fix = 1, V = 0.25 * (I + J), n = 3), G = list (G1 = list (V = diag (4), n = 4), G2 = list (V = diag (4), n = 4)))
data = Data,
pedigree= Apistree ,
nitt = 600000,
burnin = 100000,
thin = 100)
with I = J = 3 (since they are 4 categories). to implement this model throws me the following error:
V is the wrong dimension for Some priorG / priorR elements.
I have little experience in the use of R or less of MCMCglmm, so if you can help me in the structure of the model, as in the prior adjustment would be very nice.
Hi Oscar,
Have you solved this problem? I have been away for a while. If you haven’t please let me know. Otherwise, R-Sig-Mixed-Models mailing list is a great place to post questions about MCMCglmm as the author hangs around there and answers questions. Best.
Chris
https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
Hi Chris,
Do you have an email I could talk with you through? I have a couple of stupid questions about your profile analysis R program.
desja004 AT umn DOT edu
I am not extremely knowledgeable about profile analysis but I’d be happy to help or add any thing that you think might be interesting (or if you found a bug).
Hello, I’m a PhD at UW-Madison, and now dealing with a project that needs the idea of Zero-inflation Poisson with lasso penalty. Hope to see your wonderful work soon!
Hi,
Your project sounds interesting. I’ll post soon about both my work with zero-inflated and overdispersed models and penalized regression.