0

I am carrying out an split experiment with microbiology. Totally 3 blocks I set: A,B,C; each block contains 2 replicates; each replicate contains 2 species. I want to test if the density ratio of these 2 species change with time.

I write this code:

y19 <- cbind(data19$density.E, data19$density.P)
model19 <- glmer(y19 ~ time + (1|block), binomial, data = data19)
summary(model19)

it works, but show warning:

boundary (singular) fit: see help('isSingular') Warning messages: 1: Some predictor variables are on very different scales: consider rescaling 2: In eval(family$initialize, rho) : non-integer counts in a binomial glm!

I learned that glm function can be also used, but I dont know the format of split experiments in glm, I can only write:

model19 <- glm(y19  ~ time * block, binomial, data = data19)

what the right format of glm of my question

neilfws
  • 32,751
  • 5
  • 50
  • 63
Scrooge Mike
  • 17
  • 3
  • Why didn't you try rescaling and why binomial instead of poisson? What is the target variable? – Quality Feb 17 '23 at 04:07
  • 1
    `y19` needs to be 0's or 1's for a binomial glm() to work. – MrSwaggins Feb 17 '23 at 04:34
  • @Quality the density are already transformed, and if I use poisson, it shows error with dimension mismatch. but glmer model also has warning: 1.boundary (singular) fit: see help('isSingular'). and 2. In eval(family$initialize, rho) : non-integer counts in a binomial glm. I dont know if these warning matter? – Scrooge Mike Feb 17 '23 at 05:58
  • @MrSwaggins ok, thank you. do you know what the format to write for split experiments in glm? – Scrooge Mike Feb 17 '23 at 06:07

0 Answers0