Replaced deprecated lgamma with loggamma

src/likelihoods/binomial.jl

@@ -15,7 +15,7 @@ end
 
 function log_dens(binomial::BinLik, f::AbstractVector, y::Vector{Int})
     θ = @. exp(f) / (1 + exp(f))
-    return Float64[lgamma(binomial.n+1.0) - lgamma(yi+1.0) - lgamma(binomial.n-yi+1.0) + yi*log(θi) + (binomial.n-yi)*log(1-θi) for (θi,yi) in zip(θ,y)]
+    return Float64[loggamma(binomial.n+1.0) - loggamma(yi+1.0) - loggamma(binomial.n-yi+1.0) + yi*log(θi) + (binomial.n-yi)*log(1-θi) for (θi,yi) in zip(θ,y)]
 end
 
 function dlog_dens_df(binomial::BinLik, f::AbstractVector, y::Vector{Int})

src/likelihoods/poisson.jl

@@ -12,7 +12,7 @@ struct PoisLik <: Likelihood end
 #log of probability density
 function log_dens(poisson::PoisLik, f::AbstractVector, y::Vector{Int})
     #where we exponentiate for positivity f = exp(fi)
-    return y.*f - exp.(f) - lgamma.(1.0 .+ y)
+    return y.*f - exp.(f) - loggamma.(1.0 .+ y)
 end
 
 #derivative of pdf wrt latent function

src/likelihoods/studentT.jl

@@ -29,7 +29,7 @@ end
 function log_dens(studentT::StuTLik, f::AbstractVector, y::AbstractVector)
     ν = studentT.ν
     σ = studentT.σ
-    c = lgamma(0.5*(ν+1)) - lgamma(0.5*ν) - 0.5*log(pi*ν) - log(σ)
+    c = loggamma(0.5*(ν+1)) - loggamma(0.5*ν) - 0.5*log(pi*ν) - log(σ)
     return [c - (0.5*(ν+1))*log(1+(1/ν)*((yi-fi)/σ)^2) for (fi,yi) in zip(f,y)]
 end