From 2fb0d614c9728f77f40156b90ab5d2b9441d9602 Mon Sep 17 00:00:00 2001
From: "Anthony D. Blaom" <anthony.blaom@gmail.com>
Date: Tue, 27 Feb 2024 11:26:53 +1300
Subject: [PATCH 1/4] add tests for categorical features

---
 test/runtests.jl | 28 ++++++++++++++++++++++++++--
 1 file changed, 26 insertions(+), 2 deletions(-)

diff --git a/test/runtests.jl b/test/runtests.jl
index 32c4df3..4a70962 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -8,6 +8,7 @@ using MLJGLMInterface
 using GLM: coeftable
 import GLM
 import MLJTestInterface
+using Tables
 
 using  Distributions: Normal, Poisson, Uniform
 import StableRNGs
@@ -15,12 +16,33 @@ using Tables
 
 expit(X) = 1 ./ (1 .+ exp.(-X))
 
-# TODO: Add more datasets to the following generic interface tests after #45 is merged
+# synthesize small data sets with mixed features:
+
+n = 100
+X_regression, y_regression = MLJBase.make_regression(n, 3)
+outlook = categorical(rand(["sunny", "overcast", "rainy"], n))
+temperature = categorical(
+    rand(["cold", "mild", "hot"], n),
+    ordered=true,
+    levels=["cold", "mild", "hot"],
+    )
+X = merge(
+    Tables.columntable(X_regression),
+    (; outlook, temperature),
+)
+y_binary = (temperature .== "mild" .|| outlook .== "sunny") |> categorical
+y_count = map(X.x1) do x
+    floor(Int, 10*abs(x))
+end
+mixed_binary = (X, y_binary)
+mixed_count = (X, y_count)
+mixed_regression = (X, y_regression)
 
 @testset "generic interface tests" begin
     @testset "LinearRegressor" begin
         for data in [
-                        MLJTestInterface.make_regression(),
+            MLJTestInterface.make_regression(),
+            mixed_regression,
             ]
             failures, summary = MLJTestInterface.test(
                 [LinearRegressor,],
@@ -36,6 +58,7 @@ expit(X) = 1 ./ (1 .+ exp.(-X))
     @testset "LinearCountRegressor" begin
         for data in [
             MLJTestInterface.make_count(),
+            mixed_count,
             ]
             failures, summary = MLJTestInterface.test(
                 [LinearCountRegressor,],
@@ -51,6 +74,7 @@ expit(X) = 1 ./ (1 .+ exp.(-X))
     @testset "LinearBinaryClassifier" begin
         for data in [
             MLJTestInterface.make_binary(),
+            mixed_binary,
             ]
             failures, summary = MLJTestInterface.test(
                 [LinearBinaryClassifier,],

From 9d32af8c2fa4e7060b75cbb39ed1e0a47ce70e4a Mon Sep 17 00:00:00 2001
From: "Anthony D. Blaom" <anthony.blaom@gmail.com>
Date: Wed, 28 Feb 2024 17:15:37 +1300
Subject: [PATCH 2/4] try to fix problem with new test in julia 1.6

---
 test/runtests.jl | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/test/runtests.jl b/test/runtests.jl
index 4a70962..98320e1 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -30,7 +30,7 @@ X = merge(
     Tables.columntable(X_regression),
     (; outlook, temperature),
 )
-y_binary = (temperature .== "mild" .|| outlook .== "sunny") |> categorical
+y_binary = categorical(temperature .== "mild" .|| outlook .== "sunny")
 y_count = map(X.x1) do x
     floor(Int, 10*abs(x))
 end

From 048c55b27176bd9db9994e361cdf4f93cada0e1d Mon Sep 17 00:00:00 2001
From: "Anthony D. Blaom" <anthony.blaom@gmail.com>
Date: Wed, 28 Feb 2024 17:20:06 +1300
Subject: [PATCH 3/4] try again

---
 test/runtests.jl | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/test/runtests.jl b/test/runtests.jl
index 98320e1..351a895 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -30,7 +30,7 @@ X = merge(
     Tables.columntable(X_regression),
     (; outlook, temperature),
 )
-y_binary = categorical(temperature .== "mild" .|| outlook .== "sunny")
+y_binary = categorical(temperature .== "mild" .| outlook .== "sunny")
 y_count = map(X.x1) do x
     floor(Int, 10*abs(x))
 end

From 886deee0e5b3ac25f365403a738ed983e69ef7f1 Mon Sep 17 00:00:00 2001
From: "Anthony D. Blaom" <anthony.blaom@gmail.com>
Date: Wed, 28 Feb 2024 18:07:05 +1300
Subject: [PATCH 4/4] and again...

---
 test/runtests.jl | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/test/runtests.jl b/test/runtests.jl
index 351a895..8d475dc 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -30,7 +30,7 @@ X = merge(
     Tables.columntable(X_regression),
     (; outlook, temperature),
 )
-y_binary = categorical(temperature .== "mild" .| outlook .== "sunny")
+y_binary = categorical((temperature .== "mild") .| (outlook .== "sunny"))
 y_count = map(X.x1) do x
     floor(Int, 10*abs(x))
 end