1

I am trying to execute gamlss's zero-inflated beta regression model as below> It works fine with variable V255 but through an error for simialr variable V256 as 

 Error in glim.fit(f = nu.object, X = nu.X, y = y, w = w, fv = nu, os = nu.offset,  : 
      object 'fit' not foundim.fit. Object 'fit' not found.

Any suggestions to fix this issue ? Code and model.

library(tidyverse)
library(gamlss)

ddf<-as_tibble(structure(c(0, 0.250601410865784, 0, 0, 0, 0, 0, 0, 0, 0, 0, 
                            0, 0, 0, 0.0859533622860909, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 
                            0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.070460669696331, 
                            0, 0, 0, 0.0223423503339291, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 
                            0, 0, 0, 0, 0, 0, 0, 0, 0, 0.249115347862244, 0, 0, 0, 0, 0, 
                            0, 0, 0, 0, 0, 0, 0, 0.135218366980553, 0, 0, 0, 0, 0, 0, 0, 
                            0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.069370836019516, 
                            0, 0, 0, 0.0225795395672321, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 
                            0, 0, 0, 0, 0, 0, 0, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 
                            13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 
                            29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 
                            45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 
                            61, 62, 63, 64, 65), .Dim = c(65L, 3L), .Dimnames = list(NULL, 
                                                                                     c("V255", "V256", "year"))))
ddf
#Model for variable V255, it works 
model1<-gamlss(ddf$V255~ddf$year, sigma.fo=~ddf$year, nu.fo=~ddf$year,tau.fo=~ddf$year, family=BEINF)
model1
#Model for varable V256; it through an error
model2<-gamlss(ddf$V256~ddf$year, sigma.fo=~ddf$year, nu.fo=~ddf$year,tau.fo=~ddf$year, family=BEINF)
Lily Nature
  • 613
  • 7
  • 18
  • Hm, for me, your examples (both models) run fine. – Daniel Nov 26 '18 at 07:59
  • I don't get an error, but all the estimates in the model2 object are NULL. After looking at the data with `plot` it appears these are both rather pathological datasets. So I think you have just discovered one of the boundaries of the feasible region for estimation. – IRTFM Nov 26 '18 at 08:01
  • Thank you, to my surprise i am getting the same error when tried with another computer too. Any fix ? – Lily Nature Nov 26 '18 at 17:31

1 Answers1

0

method=GC() fix the problem.

model2<-gamlss(ddf$V256~ddf$year, sigma.fo=~ddf$year, nu.fo=~ddf$year,tau.fo=~ddf$year, family=BEINF , method=GC())
Suraj Sukale
  • 1,778
  • 1
  • 12
  • 19
Tito
  • 1
  • 1