N-mixture models in Python

Question

I'm new to Python and am having trouble translating a model I wrote in R into Python language. If anyone has any suggestions on resources or code examples that might help I would greatly appreciate it. I've seen some snippets of code and text in help files, etc. but none are quite annotated or specific enough for a newbie to Python. The following model is an N-mixture abundance model modeled after Royle (2004):N-mixture models for estimating population size from spatially replicated counts. Basically it describes a Poisson/Binomial mixture model where-in the Where Z_i is wetland-level abundance and is treated as a random variable with a Poisson distribution. The observed abundance of broods (yij) on site i and during visit j then follows a binomial distribution with index parameter Z_i and success parameter p_ij. Abundance 〖(λ〗_i) is modeled through a log link as a function of a covariates and detection probabilities are modeled through a logit link as a function of b covariates.

model {

## Priors

a0 ~ dunif(-5, 5)
a1~ dunif(-5, 5)
a2 ~ dunif(-5, 5)
a3~ dunif(-5, 5)
b0 ~ dunif(-5, 5)
b1~ dunif(-5, 5)
b2~ dunif(-5, 5)

## Model

# State process
for(i in 1:5175) {
logit(psi[i]) <- min(max(a0 + a1*wetarea[i] + 
    a2*percentcover[i] + 
    a3*(year[i]), -99), 99)

Z[i] ~ dbern(psi[i])

# Detection process
for(j in 1:3) {
    logit(p[i, j]) <- b0 + b1*emergentcover[i, j] + 
        b2*time[i]
    y[i, j] ~ dbin(p[i, j], Z[i])
    }
}


## Derived parameters   
Zsum <- sum(Z[])    # Number of sites occupied
PAO <- Zsum / 100   # Proportion of sites occupied (aka PAO)

}

Thanks in advance for any help/suggestions

Answer 1

There are several examples on the PyMC wiki, including occupancy and mixture models. I would look at those; your model looks pretty straightforward.