[R] Help on VGAM package(missing value where TRUE/FALSE needed )

chamilka chamilka.john at gmail.com
Tue Jul 31 22:17:57 CEST 2012


I am using  VGAM R package for my research works. 
In my study, I am trying to simulate 1000 datasets from Beta binomial
distribution for a given set of (n,prob,size,rho)
and then I need to estimate the parameters prob and rho of the simulated
1000 datasets and need to store the fitted coefficients in 1000 by 2 data
frame. For that purpose, by referring to VGAM user manual, I tried the
following R coded..
*
bdata = data.frame(N = 5, mu = 0.2, rho = 0.9) #define the number of trials,
mean,rho 
bdata = cbind(bdata,replicate(1000,
rbetabinom(n=10,prob=0.2,size=5,rho=0.9))) #concatenate the above with a
simulated 1000 beta binomial datasets

ab<-c() #defining a null object for the purpose of storage  
#Below I am running a for loop by treating each column of bdata as a
response(from 4th) to estimate the parameters
for(i in 4:100){
  fit = vglm(cbind(bdata[,i], N-bdata[,i]) ~ 1, betabinomial, bdata, trace =
TRUE)
  ab<-rbind(ab,Coef(fit))
} *


The code is running, but I am getting an error of below form
/Error in if (any(near0.5 <- (abs(theta - 0.5) < 0.000125))) temp2[near0.5]
= log(theta[near0.5]/(1 -  : 
  missing value where TRUE/FALSE needed
/
Please help me on this error. 

Thanks in advance



--
View this message in context: http://r.789695.n4.nabble.com/Help-on-VGAM-package-missing-value-where-TRUE-FALSE-needed-tp4638595.html
Sent from the R help mailing list archive at Nabble.com.



More information about the R-help mailing list