Automatic JuliaFormatter.jl run

docs/make.jl

@@ -13,7 +13,7 @@ makedocs(;
     ),
     pages = [
         "Home" => "index.md",
-#        "Example: Gaussian Mixture" => "example_1.md",
+        #        "Example: Gaussian Mixture" => "example_1.md",
         #"Reference" => "reference.md",
     ],
 )

examples/example_n1.jl

@@ -45,7 +45,7 @@ res = sample(
 
 # You can also use Sequential Monte Carlo (SMC) to infer posterior parameters:
 
-ressmc = smc(prior, cost, nparticles=500, epstol=0.01)
+ressmc = smc(prior, cost, nparticles = 500, epstol = 0.01)
 
 @show ressmc
 

examples/example_n2.jl

@@ -9,8 +9,8 @@ function model(P, N)
 
     d1 = @. r1 * σ_1 + μ_1
     d2 = @. r1 * σ_2 + μ_2
-    ps = @. (1 + sign(r2 - prob))/2
-    @. (d1+ps*(d2-d1))
+    ps = @. (1 + sign(r2 - prob)) / 2
+    @. (d1 + ps * (d2 - d1))
 
 end
 
@@ -56,5 +56,5 @@ approx_density = ApproxPosterior(prior, D, 0.032)
 @show res
 
 # In this case, it is best to apply SMC, as it leads to tighter CI's and lower computational costs
-@time res = smc(prior, D, verbose=false, nparticles=100, alpha=0.95)
+@time res = smc(prior, D, verbose = false, nparticles = 100, alpha = 0.95)
 @show res.P

src/KissABC.jl

@@ -55,8 +55,9 @@ function AbstractMCMC.step(
             particles[i] = op(float, unconditional_sample(rng, model))
             logdensity[i] = loglike(model, push_p(model, particles[i]))
             retrys -= 1
-            retrys < 0 &&
-                error("Prior leads to ∞ costs too often, tune the prior or increase `retry_sampling`.")
+            retrys < 0 && error(
+                "Prior leads to ∞ costs too often, tune the prior or increase `retry_sampling`.",
+            )
         end
     end