Rename GAM.jl to GeneralizedAdditiveModels.jl

Project.toml

@@ -1,4 +1,4 @@
-name = "GAM"
+name = "GeneralizedAdditiveModels"
 uuid = "cc454e9f-ce0f-441e-b193-468e31ddef4b"
 authors = ["Trent Henderson <trent.henderson1@outlook.com>"]
 version = "0.1.0"

README.md

@@ -1,4 +1,4 @@
-# GAM.jl
+# GeneralizedAdditiveModels.jl 
 Fit, evaluate, and visualise generalised additive models (GAMs) in native Julia
 
 ## Motivation
@@ -7,7 +7,7 @@ Fit, evaluate, and visualise generalised additive models (GAMs) in native Julia
 
 ## Usage
 
-The basic interface to `GAM.jl` is the `gam` function, which is as easy as:
+The basic interface to `GeneralizedAdditiveModels.jl` is the `gam` function, which is as easy as:
 
 ```{julia}
     mod = gam("Volume ~ s(Girth, k=10, degree=3) + s(Height, k=10, degree=3)", df)
@@ -28,12 +28,14 @@ Note that currently, the following families are supported:
 * `Normal`
 * `Poisson`
 * `Gamma`
+* `Bernoulli`
 
 And the following link functions:
 
 * `Identity`
 * `Log`
+* `Logit`
 
 ## Development notes
 
-`GAM.jl` is very much in active development. Please check back for updates and new features or feel free to contribute yourself! The project to-date has been a collaboration between [Trent Henderson](https://github.com/hendersontrent) and [Mason Yahr](https://github.com/yahrMason).
+`GeneralizedAdditiveModels.jl` is very much in active development. Please check back for updates and new features or feel free to contribute yourself! The project to-date has been a collaboration between [Trent Henderson](https://github.com/hendersontrent) and [Mason Yahr](https://github.com/yahrMason).

docs/Project.toml

@@ -1,5 +1,5 @@
 [deps]
 Documenter = "e30172f5-a6a5-5a46-863b-614d45cd2de4"
-GAM = "cc454e9f-ce0f-441e-b193-468e31ddef4b"
+GeneralizedAdditiveModels = "cc454e9f-ce0f-441e-b193-468e31ddef4b"
 LiveServer = "16fef848-5104-11e9-1b77-fb7a48bbb589"
 RDatasets = "ce6b1742-4840-55fa-b093-852dadbb1d8b"

docs/make.jl

@@ -3,14 +3,14 @@ using Documenter
 
 # Make sure the package loads when building from docs/
 push!(LOAD_PATH, joinpath(@__DIR__, "..", "src"))
-using GAM
+using GeneralizedAdditiveModels  
 
 # Doctesting
 DocMeta.setdocmeta!(GAM, :DocTestSetup, :(using GAM); recursive=true)
 
 makedocs(
-    sitename = "GAM.jl",
-    modules  = [GAM],
+    sitename = "GeneralizedAdditiveModels.jl",
+    modules  = [GeneralizedAdditiveModels],
     pages    = [
         "Home" => "index.md",
         "API Reference" => "api_reference.md",
@@ -22,4 +22,4 @@ makedocs(
 )
 
 # for later
-# deploydocs(repo = "github.com/hendersontrent/GAM.jl.git", devbranch = "main")
+# deploydocs(repo = "github.com/hendersontrent/GeneralizedAdditiveModels.jl.git", devbranch = "main")

docs/src/api_reference.md

@@ -1,9 +1,9 @@
 # API Reference
 
 ```@autodocs
-Modules   = [GAM]
+Modules   = [GeneralizedAdditiveModels]
 Recursive = true
 Public    = true
 Private   = false
 Order     = [:module, :type, :function, :macro, :constant]
-```
+```

docs/src/index.md

@@ -33,7 +33,7 @@ To install the package, clone it directly from the repository:
 
 ```julia
 using Pkg
-Pkg.add(url="https://github.com/hendersontrent/GAM.jl")
+Pkg.add(url="https://github.com/hendersontrent/GeneralizedAdditiveModels.jl")
 ```
 
 ---
@@ -44,7 +44,7 @@ Fitting a GAM in GeneralizedAdditiveModels.jl is quick and easy. The syntax of t
 
 ```julia
 using RDatasets
-using GAM
+using GeneralizedAdditiveModels
 
 df = dataset("datasets", "trees");
 
@@ -55,7 +55,7 @@ mod = gam("Volume ~ s(Girth, k=10, degree=3) + s(Height, k=10, degree=3)", df)
 
 **GeneralizedAdditiveModels.jl** is under active development, and contributions are very welcome!
 
-- If you’ve found a bug or want to propose a feature, please [open an issue](https://github.com/hendersontrent/GAM.jl/issues).
+- If you’ve found a bug or want to propose a feature, please [open an issue](https://github.com/hendersontrent/GeneralizedAdditiveModels.jl/issues).
 - If your idea gets positive feedback, feel free to submit a pull request.
 - If you’re unsure where to start, you can also browse the open issues and pick one that interests you.
 

src/GAM.jl → src/GeneralizedAdditiveModels.jl

@@ -1,4 +1,4 @@
-module GAM
+module GeneralizedAdditiveModels
 
 using Distributions, GLM, Optim, BSplines, LinearAlgebra, DataFrames, Plots, Optim
 

src/alpha.jl

@@ -4,7 +4,7 @@ Calculate alpha.
 
 Usage:
 ```julia-repl
-alpha((y, mu, Dist, Link)
+alpha(y, mu, Dist, Link)
 ```
 Arguments:
 - `y` : `Vector` containing the response variable.

test/runtests.jl

@@ -1,11 +1,11 @@
-using GAM
+using GeneralizedAdditiveModels
 using Test
 using RDatasets, DataFrames, Plots
 using Distributions, Random, Statistics, LinearAlgebra
 
 #-------------------- Run tests -----------------
 
-@testset "GAM.jl Test Suite" begin
+@testset "GeneralizedAdditiveModels.jl" begin
 
     # ===========================
     # Core Gaussian GAM Tests
@@ -44,7 +44,7 @@ using Distributions, Random, Statistics, LinearAlgebra
 
             # Test partial predictions for each smooth
             for i in 1:length(mod.x)
-                partial = GAM.PredictPartial(mod, i)
+                partial = GeneralizedAdditiveModels.PredictPartial(mod, i)
                 @test length(partial) == length(mod.x[i])
                 @test !any(isnan.(partial))
                 @test !any(isinf.(partial))
@@ -202,7 +202,7 @@ using Distributions, Random, Statistics, LinearAlgebra
 
             # Linear term should have minimal degrees of freedom
             # (This assumes Height is the second term)
-            partial_height = GAM.PredictPartial(mod, 2)
+            partial_height = GeneralizedAdditiveModels.PredictPartial(mod, 2)
             @test length(unique(diff(partial_height[sortperm(df.Height)]))) < 5  # Should be nearly constant slope
         end
 
@@ -229,12 +229,12 @@ using Distributions, Random, Statistics, LinearAlgebra
             @test all(0 .<= mod.Fitted .<= 1)
 
             # Check that linear effect is captured
-            partial_linear = GAM.PredictPartial(mod, 1)
+            partial_linear = GeneralizedAdditiveModels.PredictPartial(mod, 1)
             linear_coef = cov(partial_linear, x_linear) / var(x_linear)
             @test abs(linear_coef - beta_linear) < 0.5  # Should be close to true value
 
             # Check that smooth captures nonlinearity
-            partial_smooth = GAM.PredictPartial(mod, 2)
+            partial_smooth = GeneralizedAdditiveModels.PredictPartial(mod, 2)
             smooth_sorted = partial_smooth[sortperm(x_smooth)]
 
             # Should have multiple turning points (capturing the sine wave)
@@ -295,16 +295,16 @@ using Distributions, Random, Statistics, LinearAlgebra
         x = randn(100)
 
         @testset "Uniform basis" begin
-            basis = GAM.BuildUniformBasis(x, 10, 3)
+            basis = GeneralizedAdditiveModels.BuildUniformBasis(x, 10, 3)
             @test length(basis.breakpoints) == 10
             @test basis.order == 3
             @test minimum(basis.breakpoints) ≈ minimum(x)
             @test maximum(basis.breakpoints) ≈ maximum(x)
         end
 
         @testset "Basis matrix properties" begin
-            basis = GAM.BuildUniformBasis(x, 8, 3)
-            X = GAM.BuildBasisMatrix(basis, x)
+            basis = GeneralizedAdditiveModels.BuildUniformBasis(x, 8, 3)
+            X = GeneralizedAdditiveModels.BuildBasisMatrix(basis, x)
 
             @test size(X, 1) == length(x)
             @test size(X, 2) == length(basis)
@@ -313,8 +313,8 @@ using Distributions, Random, Statistics, LinearAlgebra
         end
 
         @testset "Penalty matrix" begin
-            basis = GAM.BuildUniformBasis(x, 10, 3)
-            D = GAM.BuildDifferenceMatrix(basis)
+            basis = GeneralizedAdditiveModels.BuildUniformBasis(x, 10, 3)
+            D = GeneralizedAdditiveModels.BuildDifferenceMatrix(basis)
 
             @test size(D, 2) == length(basis)
             @test size(D, 1) == length(basis) - 2  # Second differences