feat(turing): juliaformatter blue style

flake.nix

@@ -58,6 +58,27 @@
                 files = "\\.typ$";
                 language = "rust";
               };
+              julia-formatter = {
+                enable = true;
+                name = "format julia code";
+                entry = ''
+                  ${pkgs.julia-bin}/bin/julia -e '
+                  using Pkg
+                  Pkg.activate(".")
+                  using JuliaFormatter
+                  format(ARGS)
+                  out = Cmd(`git diff --name-only`) |> read |> String
+                  if out == ""
+                      exit(0)
+                  else
+                      @error "Some files have been formatted !!!"
+                      write(stdout, out)
+                      exit(1)
+                  end'
+                '';
+                files = "\\.jl$";
+                language = "system";
+              };
             };
             settings = {
               treefmt.package = treefmtEval.config.build.wrapper;

turing/.JuliaFormatter.toml

@@ -0,0 +1 @@
+style = "blue"

turing/02-linear_regression-kidiq.jl

@@ -13,11 +13,11 @@ seed!(123)
 kidiq = CSV.read("datasets/kidiq.csv", DataFrame)
 
 # define data matrix X and standardize
-X = select(kidiq, Not(:kid_score)) |> Matrix
+X = Matrix(select(kidiq, Not(:kid_score)))
 X = standardize(ZScoreTransform, X; dims=1)
 
 # define dependent variable y and standardize
-y = kidiq[:, :kid_score] |> float
+y = float(kidiq[:, :kid_score])
 y = standardize(ZScoreTransform, y; dims=1)
 
 # define the model
@@ -61,8 +61,12 @@ model_qr = linear_regression(Q_ast, y)
 chn_qr = sample(model_qr, NUTS(1_000, 0.8), MCMCThreads(), 1_000, 4)
 
 # reconstruct β back from the Q_ast scale into X scale
-betas = mapslices(x -> R_ast^-1 * x, chn_qr[:, namesingroup(chn_qr, :β), :].value.data, dims=[2])
-chain_beta = setrange(Chains(betas, ["real_β[$i]" for i in 1:size(Q_ast, 2)]), 1_001:1:2_000)
+betas = mapslices(
+    x -> R_ast^-1 * x, chn_qr[:, namesingroup(chn_qr, :β), :].value.data; dims=[2]
+)
+chain_beta = setrange(
+    Chains(betas, ["real_β[$i]" for i in 1:size(Q_ast, 2)]), 1_001:1:2_000
+)
 chn_qr_reconstructed = hcat(chain_beta, chn_qr)
 println(DataFrame(summarystats(chn_qr_reconstructed)))
 

turing/03-logistic_regression-wells.jl

@@ -14,7 +14,7 @@ seed!(123)
 wells = CSV.read("datasets/wells.csv", DataFrame)
 
 # define data matrix X and standardize
-X = select(wells, Not(:switch)) |> Matrix
+X = Matrix(select(wells, Not(:switch)))
 X = standardize(ZScoreTransform, X; dims=1)
 
 # define dependent variable y

turing/04-ordinal_regression-esoph.jl

@@ -25,7 +25,7 @@ transform!(
         x -> categorical(x; levels=["0-39g/day", "40-79", "80-119", "120+"], ordered=true),
     :tobgp =>
         x -> categorical(x; levels=["0-9g/day", "10-19", "20-29", "30+"], ordered=true);
-    renamecols=false
+    renamecols=false,
 )
 transform!(esoph, [:agegp, :alcgp, :tobgp] .=> ByRow(levelcode); renamecols=false)
 

turing/05-poisson_regression-roaches.jl

@@ -14,7 +14,7 @@ seed!(123)
 roaches = CSV.read("datasets/roaches.csv", DataFrame)
 
 # define data matrix X and standardize
-X = select(roaches, Not(:y)) |> Matrix
+X = Matrix(select(roaches, Not(:y)))
 X = standardize(ZScoreTransform, X; dims=1)
 
 # define dependent variable y

turing/06-robust_linear_regression-duncan.jl

@@ -13,11 +13,11 @@ seed!(123)
 duncan = CSV.read("datasets/duncan.csv", DataFrame)
 
 # define data matrix X and standardize
-X = select(duncan, [:income, :education]) |> Matrix |> float
+X = float(Matrix(select(duncan, [:income, :education])))
 X = standardize(ZScoreTransform, X; dims=1)
 
 # define dependent variable y and standardize
-y = duncan[:, :prestige] |> float
+y = float(duncan[:, :prestige])
 y = standardize(ZScoreTransform, y; dims=1)
 
 # define the model

turing/07-robust_beta_binomial_regression-wells.jl

@@ -17,7 +17,7 @@ seed!(123)
 wells = CSV.read("datasets/wells.csv", DataFrame)
 
 # define data matrix X and standardize
-X = select(wells, Not(:switch)) |> Matrix
+X = Matrix(select(wells, Not(:switch)))
 X = standardize(ZScoreTransform, X; dims=1)
 
 # define dependent variable y

turing/08-robust_robit_regression-wells.jl

@@ -14,7 +14,7 @@ seed!(123)
 wells = CSV.read("datasets/wells.csv", DataFrame)
 
 # define data matrix X and standardize
-X = select(wells, Not(:switch)) |> Matrix
+X = Matrix(select(wells, Not(:switch)))
 X = standardize(ZScoreTransform, X; dims=1)
 
 # define dependent variable y

turing/09-robust_negative_binomial_regression-roaches.jl

@@ -14,7 +14,7 @@ seed!(123)
 roaches = CSV.read("datasets/roaches.csv", DataFrame)
 
 # define data matrix X and standardize
-X = select(roaches, Not(:y)) |> Matrix
+X = Matrix(select(roaches, Not(:y)))
 X = standardize(ZScoreTransform, X; dims=1)
 
 # define dependent variable y

turing/10-robust_zero_inflated_regression-negative_binomial-roaches.jl

@@ -13,7 +13,7 @@ seed!(123)
 roaches = CSV.read("datasets/roaches.csv", DataFrame)
 
 # define data matrix X and standardize
-X = select(roaches, Not(:y)) |> Matrix
+X = Matrix(select(roaches, Not(:y)))
 X = standardize(ZScoreTransform, X; dims=1)
 
 # define dependent variable y
@@ -28,7 +28,9 @@ function NegativeBinomial2(μ, ϕ)
 end
 
 # define the model
-@model function zero_inflated_negative_binomial_regression(X, y; predictors=size(X, 2), N=size(X, 1))
+@model function zero_inflated_negative_binomial_regression(
+    X, y; predictors=size(X, 2), N=size(X, 1)
+)
     # priors
     α ~ TDist(3) * 2.5
     β ~ filldist(TDist(3) * 2.5, predictors)
@@ -40,11 +42,11 @@ end
     for n in 1:N
         if y[n] == 0
             Turing.@addlogprob! logpdf(Bernoulli(γ), 0) +
-                                logpdf(Bernoulli(γ), 1) +
-                                logpdf(NegativeBinomial2(exp(α + X[n, :] ⋅ β), ϕ), y[n])
+                logpdf(Bernoulli(γ), 1) +
+                logpdf(NegativeBinomial2(exp(α + X[n, :] ⋅ β), ϕ), y[n])
         else
             Turing.@addlogprob! logpdf(Bernoulli(γ), 0) +
-                                logpdf(NegativeBinomial2(exp(α + X[n, :] ⋅ β), ϕ), y[n])
+                logpdf(NegativeBinomial2(exp(α + X[n, :] ⋅ β), ϕ), y[n])
         end
     end
     return (; y, α, β, γ, ϕ)

turing/10-robust_zero_inflated_regression-poisson-roaches.jl

@@ -13,7 +13,7 @@ seed!(123)
 roaches = CSV.read("datasets/roaches.csv", DataFrame)
 
 # define data matrix X and standardize
-X = select(roaches, Not(:y)) |> Matrix
+X = Matrix(select(roaches, Not(:y)))
 X = standardize(ZScoreTransform, X; dims=1)
 
 # define dependent variable y
@@ -30,11 +30,11 @@ y = roaches[:, :y]
     for n in 1:N
         if y[n] == 0
             Turing.@addlogprob! logpdf(Bernoulli(γ), 0) +
-                                logpdf(Bernoulli(γ), 1) +
-                                logpdf(LogPoisson(α + X[n, :] ⋅ β), y[n])
+                logpdf(Bernoulli(γ), 1) +
+                logpdf(LogPoisson(α + X[n, :] ⋅ β), y[n])
         else
             Turing.@addlogprob! logpdf(Bernoulli(γ), 0) +
-                                logpdf(LogPoisson(α + X[n, :] ⋅ β), y[n])
+                logpdf(LogPoisson(α + X[n, :] ⋅ β), y[n])
         end
     end
     return (; y, α, β, γ)

turing/11-sparse_horseshoe.jl

@@ -13,11 +13,11 @@ seed!(123)
 df = CSV.read("datasets/sparse_regression.csv", DataFrame)
 
 # define data matrix X and standardize
-X = select(df, Not(:y)) |> Matrix |> float
+X = float(Matrix(select(df, Not(:y))))
 X = standardize(ZScoreTransform, X; dims=1)
 
 # define dependent variable y and standardize
-y = df[:, :y] |> float
+y = float(df[:, :y])
 y = standardize(ZScoreTransform, y; dims=1)
 
 # define the model

turing/12-sparse_horseshoe_p.jl

@@ -13,11 +13,11 @@ seed!(123)
 df = CSV.read("datasets/sparse_regression.csv", DataFrame)
 
 # define data matrix X and standardize
-X = select(df, Not(:y)) |> Matrix |> float
+X = float(Matrix(select(df, Not(:y))))
 X = standardize(ZScoreTransform, X; dims=1)
 
 # define dependent variable y and standardize
-y = df[:, :y] |> float
+y = float(df[:, :y])
 y = standardize(ZScoreTransform, y; dims=1)
 
 # define the model

turing/13-sparse_finnish_horseshoe.jl

@@ -13,23 +13,16 @@ seed!(123)
 df = CSV.read("datasets/sparse_regression.csv", DataFrame)
 
 # define data matrix X and standardize
-X = select(df, Not(:y)) |> Matrix |> float
+X = float(Matrix(select(df, Not(:y))))
 X = standardize(ZScoreTransform, X; dims=1)
 
 # define dependent variable y and standardize
-y = df[:, :y] |> float
+y = float(df[:, :y])
 y = standardize(ZScoreTransform, y; dims=1)
 
 # define the model
 @model function sparse_finnish_horseshoe_regression(
-    X,
-    y;
-    predictors=size(X, 2),
-    τ₀=3,
-    ν_local=1,
-    ν_global=1,
-    slab_df=4,
-    slab_scale=2,
+    X, y; predictors=size(X, 2), τ₀=3, ν_local=1, ν_global=1, slab_df=4, slab_scale=2
 )
     # priors
     α ~ TDist(3) * 2.5

turing/14-sparse_r2d2.jl

@@ -13,21 +13,16 @@ seed!(123)
 df = CSV.read("datasets/sparse_regression.csv", DataFrame)
 
 # define data matrix X and standardize
-X = select(df, Not(:y)) |> Matrix |> float
+X = float(Matrix(select(df, Not(:y))))
 X = standardize(ZScoreTransform, X; dims=1)
 
 # define dependent variable y and standardize
-y = df[:, :y] |> float
+y = float(df[:, :y])
 y = standardize(ZScoreTransform, y; dims=1)
 
 # define the model
 @model function sparse_r2d2_regression(
-    X,
-    y;
-    predictors=size(X, 2),
-    mean_R²=0.5,
-    prec_R²=2,
-    cons_D2=1,
+    X, y; predictors=size(X, 2), mean_R²=0.5, prec_R²=2, cons_D2=1
 )
     # priors
     α ~ TDist(3) * 2.5

turing/15-hierarchical_varying_intercept-cheese.jl

@@ -19,24 +19,29 @@ end
 
 # create int idx
 cheese[:, :background_int] = map(cheese[:, :background]) do b
-    b == "urban" ? 1 :
-    b == "rural" ? 2 : missing
+    if b == "urban"
+        1
+    elseif b == "rural"
+        2
+    else
+        missing
+    end
 end
 
 # define data matrix X
-X = select(cheese, Between(:cheese_A, :cheese_D)) |> Matrix
+X = Matrix(select(cheese, Between(:cheese_A, :cheese_D)))
 
 # define dependent variable y and standardize
-y = cheese[:, :y] |> float
+y = float(cheese[:, :y])
 y = standardize(ZScoreTransform, y; dims=1)
 
 # define vector of group memberships idx
 idx = cheese[:, :background_int]
 
 # define the model
-@model function varying_intercept_regression(X, idx, y;
-    predictors=size(X, 2),
-    n_gr=length(unique(idx)))
+@model function varying_intercept_regression(
+    X, idx, y; predictors=size(X, 2), n_gr=length(unique(idx))
+)
     # priors
     α ~ TDist(3) * 2.5
     β ~ filldist(TDist(3) * 2.5, predictors)

turing/15-hierarchical_varying_intercept-non_centered-cheese.jl

@@ -19,24 +19,29 @@ end
 
 # create int idx
 cheese[:, :background_int] = map(cheese[:, :background]) do b
-    b == "urban" ? 1 :
-    b == "rural" ? 2 : missing
+    if b == "urban"
+        1
+    elseif b == "rural"
+        2
+    else
+        missing
+    end
 end
 
 # define data matrix X
-X = select(cheese, Between(:cheese_A, :cheese_D)) |> Matrix
+X = Matrix(select(cheese, Between(:cheese_A, :cheese_D)))
 
 # define dependent variable y and standardize
-y = cheese[:, :y] |> float
+y = float(cheese[:, :y])
 y = standardize(ZScoreTransform, y; dims=1)
 
 # define vector of group memberships idx
 idx = cheese[:, :background_int]
 
 # define the model
-@model function varying_intercept_ncp_regression(X, idx, y;
-    predictors=size(X, 2),
-    n_gr=length(unique(idx)))
+@model function varying_intercept_ncp_regression(
+    X, idx, y; predictors=size(X, 2), n_gr=length(unique(idx))
+)
     # priors
     α ~ TDist(3) * 2.5
     β ~ filldist(TDist(3) * 2.5, predictors)

turing/16-hierarchical_varying_intercept_slope-cheese.jl

@@ -19,28 +19,32 @@ end
 
 # create int idx
 cheese[:, :background_int] = map(cheese[:, :background]) do b
-    b == "urban" ? 1 :
-    b == "rural" ? 2 : missing
+    if b == "urban"
+        1
+    elseif b == "rural"
+        2
+    else
+        missing
+    end
 end
 
 # define data matrix X
 # now we are binding a column of 1s as the first column of X
 # for the correlated intercepts
 insertcols!(cheese, :intercept => fill(1, nrow(cheese)))
-X = select(cheese, Cols(:intercept, Between(:cheese_A, :cheese_D))) |> Matrix
+X = Matrix(select(cheese, Cols(:intercept, Between(:cheese_A, :cheese_D))))
 
 # define dependent variable y and standardize
-y = cheese[:, :y] |> float
+y = float(cheese[:, :y])
 y = standardize(ZScoreTransform, y; dims=1)
 
 # define vector of group memberships idx
 idx = cheese[:, :background_int]
 
 # define the model
-@model function correlated_varying_intercept_slope_regression(X, idx, y;
-    predictors=size(X, 2),
-    N=size(X, 1),
-    n_gr=length(unique(idx)))
+@model function correlated_varying_intercept_slope_regression(
+    X, idx, y; predictors=size(X, 2), N=size(X, 1), n_gr=length(unique(idx))
+)
     # priors
     Ω ~ LKJCholesky(predictors, 2.0)
     σ ~ Exponential(1)