use model = instead of @register

Author: penelopeysm

models/dppl_gauss_unknown.jl

@@ -9,4 +9,4 @@ y = randn() .+ s * randn(n)
     y ~ filldist(Normal(m, s), N)
 end
 
-@register dppl_gauss_unknown(y)
+model = dppl_gauss_unknown(y)

models/dppl_hier_poisson.jl

@@ -24,4 +24,4 @@ lazyarray(f, x) = LazyArray(Base.broadcasted(f, x))
     y ~ arraydist(lazyarray(LogPoisson, alpha))
 end
 
-@register dppl_hier_poisson(y, x, idx, ns)
+model = dppl_hier_poisson(y, x, idx, ns)

models/dppl_high_dim_gauss.jl

@@ -2,4 +2,4 @@
     m ~ filldist(Normal(0, 1), D)
 end
 
-@register dppl_high_dim_gauss(10_000)
+model = dppl_high_dim_gauss(10_000)

models/dppl_hmm_semisup.jl

@@ -56,4 +56,4 @@ end
     DynamicPPL.@addlogprob! logsumexp(gamma)
 end
 
-@register dppl_hmm_semisup(K, T, T_unsup, w, z, u, alpha, beta)
+model = dppl_hmm_semisup(K, T, T_unsup, w, z, u, alpha, beta)

models/dppl_lda.jl

@@ -31,4 +31,4 @@ end
     DynamicPPL.@addlogprob! sum(log_phi_dot_theta[CartesianIndex.(w, doc)])
 end
 
-@register dppl_lda(k, m, w_lda, doc_lda, alpha, beta)
+model = dppl_lda(k, m, w_lda, doc_lda, alpha, beta)

models/dppl_logistic_regression.jl

@@ -18,4 +18,4 @@ lazyarray(f, x) = LazyArray(Base.broadcasted(f, x))
     y ~ arraydist(lazyarray(x -> Bernoulli(safelogistic(x)), Xt * w))
 end
 
-@register dppl_logistic_regression(X', y)
+model = dppl_logistic_regression(X', y)

models/dppl_naive_bayes.jl

@@ -24,4 +24,4 @@ labels = labels[1:N]
     image_vec ~ MvNormal(vec(m[labels, :]), I)
 end
 
-@register dppl_naive_bayes(image_vec, labels, C, D)
+model = dppl_naive_bayes(image_vec, labels, C, D)

models/dppl_sto_volatility.jl

@@ -20,4 +20,4 @@ y = map(to_num, data[1:500, 2])
     end
 end
 
-@register dppl_sto_volatility(y)
+model = dppl_sto_volatility(y)