diff --git a/test/functions.cpp b/test/functions.cpp
index 87c9a9b9e..db75efec4 100644
--- a/test/functions.cpp
+++ b/test/functions.cpp
@@ -1,156 +1,77 @@
 #include <iostream>
 #include "gtest/gtest.h"
 
-#include <llmr/style/properties.hpp>
+#include <llmr/style/function_properties.hpp>
 
 using namespace llmr;
 
-
-TEST(Function, Null) {
-    EXPECT_EQ(0.0f, llmr::functions::null(0, { }));
-    EXPECT_EQ(0.0f, llmr::functions::null(0, { 3 }));
-    EXPECT_EQ(0.0f, llmr::functions::null(0, { 3, 4 }));
-    EXPECT_EQ(0.0f, llmr::functions::null(4, { }));
-    EXPECT_EQ(0.0f, llmr::functions::null(4, { 3 }));
-    EXPECT_EQ(0.0f, llmr::functions::null(4, { 3, 4 }));
-    EXPECT_EQ(0.0f, llmr::functions::null(22, { }));
-    EXPECT_EQ(0.0f, llmr::functions::null(22, { 3 }));
-    EXPECT_EQ(0.0f, llmr::functions::null(22, { 3, 4 }));
-}
-
-
 TEST(Function, Constant) {
-    EXPECT_EQ(2.0f, llmr::functions::constant(0, { 2 }));
-    EXPECT_EQ(3.8f, llmr::functions::constant(0, { 3.8 }));
-    EXPECT_EQ(22.0f, llmr::functions::constant(0, { 22 }));
-    EXPECT_EQ(2.0f, llmr::functions::constant(4, { 2 }));
-    EXPECT_EQ(3.8f, llmr::functions::constant(4, { 3.8 }));
-    EXPECT_EQ(22.0f, llmr::functions::constant(4, { 22 }));
-    EXPECT_EQ(2.0f, llmr::functions::constant(22, { 2 }));
-    EXPECT_EQ(3.8f, llmr::functions::constant(22, { 3.8 }));
-    EXPECT_EQ(22.0f, llmr::functions::constant(22, { 22 }));
-
-    EXPECT_EQ(0.0f, llmr::functions::constant(0, { }));
-    EXPECT_EQ(0.0f, llmr::functions::constant(4, { }));
-    EXPECT_EQ(0.0f, llmr::functions::constant(22, { }));
-
-    EXPECT_EQ(4.0f, llmr::functions::constant(0, { 4.0f, 63.2f }));
-    EXPECT_EQ(4.0f, llmr::functions::constant(4, { 4.0f, 63.2f }));
-    EXPECT_EQ(4.0f, llmr::functions::constant(22, { 4.0f, 63.2f }));
+    EXPECT_EQ(2.0f, llmr::ConstantFunction<float>(2).evaluate(0));
+    EXPECT_EQ(3.8f, llmr::ConstantFunction<float>(3.8).evaluate(0));
+    EXPECT_EQ(22.0f, llmr::ConstantFunction<float>(22).evaluate(0));
+    EXPECT_EQ(2.0f, llmr::ConstantFunction<float>(2).evaluate(4));
+    EXPECT_EQ(3.8f, llmr::ConstantFunction<float>(3.8).evaluate(4));
+    EXPECT_EQ(22.0f, llmr::ConstantFunction<float>(22).evaluate(4));
+    EXPECT_EQ(2.0f, llmr::ConstantFunction<float>(2).evaluate(22));
+    EXPECT_EQ(3.8f, llmr::ConstantFunction<float>(3.8).evaluate(22));
+    EXPECT_EQ(22.0f, llmr::ConstantFunction<float>(22).evaluate(22));
 }
 
-
-TEST(Function, Min) {
-    EXPECT_EQ(0, llmr::functions::min(0, { 2 }));
-    EXPECT_EQ(0, llmr::functions::min(1.98, { 2 }));
-    EXPECT_EQ(1, llmr::functions::min(2, { 2 }));
-    EXPECT_EQ(1, llmr::functions::min(22, { 2 }));
-
-    EXPECT_EQ(0, llmr::functions::min(0, { }));
-    EXPECT_EQ(0, llmr::functions::min(1.98, { }));
-    EXPECT_EQ(0, llmr::functions::min(2, { }));
-    EXPECT_EQ(0, llmr::functions::min(22, { }));
-
-    EXPECT_EQ(0, llmr::functions::min(0, { 2, 8, 13 }));
-    EXPECT_EQ(0, llmr::functions::min(1.98, { 2, 8, 13 }));
-    EXPECT_EQ(1, llmr::functions::min(2, { 2, 8, 13 }));
-    EXPECT_EQ(1, llmr::functions::min(22, { 2, 8, 13 }));
-}
-
-
-TEST(Function, Max) {
-    EXPECT_EQ(1, llmr::functions::max(0, { 2 }));
-    EXPECT_EQ(1, llmr::functions::max(2, { 2 }));
-    EXPECT_EQ(0, llmr::functions::max(2.01, { 2 }));
-    EXPECT_EQ(0, llmr::functions::max(22, { 2 }));
-
-    EXPECT_EQ(0, llmr::functions::max(0, { }));
-    EXPECT_EQ(0, llmr::functions::max(2, { }));
-    EXPECT_EQ(0, llmr::functions::max(2.01, { }));
-    EXPECT_EQ(0, llmr::functions::max(22, { }));
-
-    EXPECT_EQ(1, llmr::functions::max(0, { 2, 8, 13 }));
-    EXPECT_EQ(1, llmr::functions::max(2, { 2, 8, 13 }));
-    EXPECT_EQ(0, llmr::functions::max(2.01, { 2, 8, 13 }));
-    EXPECT_EQ(0, llmr::functions::max(22, { 2, 8, 13 }));
-}
-
-
 TEST(Function, Stops) {
     // Explicit constant slope in fringe regions.
-    std::vector<float> slope_1 = { 0, 1.5, /**/ 6, 1.5, /**/ 8, 3, /**/ 22, 3 };
-    EXPECT_EQ(1.5, llmr::functions::stops(0, slope_1));
-    EXPECT_EQ(1.5, llmr::functions::stops(4, slope_1));
-    EXPECT_EQ(1.5, llmr::functions::stops(6, slope_1));
-    ASSERT_FLOAT_EQ(2.12132, llmr::functions::stops(7, slope_1));
-    EXPECT_EQ(3.0, llmr::functions::stops(8, slope_1));
-    EXPECT_EQ(3.0, llmr::functions::stops(9, slope_1));
-    EXPECT_EQ(3.0, llmr::functions::stops(15, slope_1));
-    EXPECT_EQ(3.0, llmr::functions::stops(22, slope_1));
+    llmr::StopsFunction<float> slope_1({ { 0, 1.5 }, { 6, 1.5 }, { 8, 3 }, { 22, 3 } });
+    EXPECT_EQ(1.5, slope_1.evaluate(0));
+    EXPECT_EQ(1.5, slope_1.evaluate(4));
+    EXPECT_EQ(1.5, slope_1.evaluate(6));
+    ASSERT_FLOAT_EQ(2.12132, slope_1.evaluate(7));
+    EXPECT_EQ(3.0, slope_1.evaluate(8));
+    EXPECT_EQ(3.0, slope_1.evaluate(9));
+    EXPECT_EQ(3.0, slope_1.evaluate(15));
+    EXPECT_EQ(3.0, slope_1.evaluate(22));
 
 
     // Test constant values in fringe regions.
-    std::vector<float> slope_2 = { 6, 1.5, /**/ 8, 3 };
-    EXPECT_EQ(1.5, llmr::functions::stops(0, slope_2));
-    EXPECT_EQ(1.5, llmr::functions::stops(4, slope_2));
-    EXPECT_EQ(1.5, llmr::functions::stops(6, slope_2));
-    ASSERT_FLOAT_EQ(2.12132, llmr::functions::stops(7, slope_2));
-    EXPECT_EQ(3.0, llmr::functions::stops(8, slope_2));
-    EXPECT_EQ(3.0, llmr::functions::stops(9, slope_2));
-    EXPECT_EQ(3.0, llmr::functions::stops(15, slope_2));
-    EXPECT_EQ(3.0, llmr::functions::stops(22, slope_2));
-
-
-    // Test uneven number of values.
-    std::vector<float> slope_3 = { 6, 1.5, 8 };
-    EXPECT_EQ(0, llmr::functions::stops(2, slope_3));
-    EXPECT_EQ(0, llmr::functions::stops(6, slope_3));
-    EXPECT_EQ(0, llmr::functions::stops(12, slope_3));
-
+    llmr::StopsFunction<float> slope_2({ { 6, 1.5 }, { 8, 3 } });
+    EXPECT_EQ(1.5, slope_2.evaluate(0));
+    EXPECT_EQ(1.5, slope_2.evaluate(4));
+    EXPECT_EQ(1.5, slope_2.evaluate(6));
+    ASSERT_FLOAT_EQ(2.12132, slope_2.evaluate(7));
+    EXPECT_EQ(3.0, slope_2.evaluate(8));
+    EXPECT_EQ(3.0, slope_2.evaluate(9));
+    EXPECT_EQ(3.0, slope_2.evaluate(15));
+    EXPECT_EQ(3.0, slope_2.evaluate(22));
 
     // Test no values.
-    std::vector<float> slope_4 = { };
-    EXPECT_EQ(1, llmr::functions::stops(2, slope_4));
-    EXPECT_EQ(1, llmr::functions::stops(6, slope_4));
-    EXPECT_EQ(1, llmr::functions::stops(12, slope_4));
+    llmr::StopsFunction<float> slope_3({});
+    EXPECT_EQ(1, slope_3.evaluate(2));
+    EXPECT_EQ(1, slope_3.evaluate(6));
+    EXPECT_EQ(1, slope_3.evaluate(12));
 }
 
 
 TEST(Function, Linear) {
-    std::vector<float> slope_1 = { /* z_base */ 4, /* val */ 7.5, /* slope */ 2, /* min */ 7.5, /* max */ 20 };
-    ASSERT_FLOAT_EQ(7.5, llmr::functions::linear(3, slope_1));
-    ASSERT_FLOAT_EQ(7.5, llmr::functions::linear(4, slope_1));
-    ASSERT_FLOAT_EQ(8.5, llmr::functions::linear(4.5, slope_1));
-    ASSERT_FLOAT_EQ(9.5, llmr::functions::linear(5, slope_1));
-    ASSERT_FLOAT_EQ(11.5, llmr::functions::linear(6, slope_1));
-    ASSERT_FLOAT_EQ(19.5, llmr::functions::linear(10, slope_1));
-    ASSERT_FLOAT_EQ(20, llmr::functions::linear(11, slope_1));
-    ASSERT_FLOAT_EQ(20, llmr::functions::linear(20, slope_1));
-
-    // Invalid number of arguments ==> 0
-    std::vector<float> slope_2 = {  };
-    ASSERT_FLOAT_EQ(0, llmr::functions::linear(3, slope_2));
-    ASSERT_FLOAT_EQ(0, llmr::functions::linear(8, slope_2));
-    ASSERT_FLOAT_EQ(0, llmr::functions::linear(10, slope_2));
+    llmr::LinearFunction<float> slope_1(/* val */ 7.5, /* z_base */ 4, /* slope */ 2, /* min */ 7.5, /* max */ 20);
+    ASSERT_FLOAT_EQ(7.5, slope_1.evaluate(3));
+    ASSERT_FLOAT_EQ(7.5, slope_1.evaluate(4));
+    ASSERT_FLOAT_EQ(8.5, slope_1.evaluate(4.5));
+    ASSERT_FLOAT_EQ(9.5, slope_1.evaluate(5));
+    ASSERT_FLOAT_EQ(11.5, slope_1.evaluate(6));
+    ASSERT_FLOAT_EQ(19.5, slope_1.evaluate(10));
+    ASSERT_FLOAT_EQ(20, slope_1.evaluate(11));
+    ASSERT_FLOAT_EQ(20, slope_1.evaluate(20));
 }
 
 
 TEST(Function, Exponential) {
-    std::vector<float> slope_1 = { /* z_base */ 4, /* val */ 7.5, /* slope */ 2, /* min */ 7.5, /* max */ 20 };
-    ASSERT_FLOAT_EQ(8.6428576, llmr::functions::exponential(3, slope_1)); // 7.5 + 1.75^(3 - 4) * 2
-    ASSERT_FLOAT_EQ(9.5, llmr::functions::exponential(4, slope_1)); // 7.5 + 1.75^(4 - 4) * 2
-    ASSERT_FLOAT_EQ(10.145751, llmr::functions::exponential(4.5, slope_1)); // 7.5 + 1.75^(4.5 - 4) * 2
-    ASSERT_FLOAT_EQ(11, llmr::functions::exponential(5, slope_1)); // 7.5 + 1.75^(5 - 4) * 2
-    ASSERT_FLOAT_EQ(13.625, llmr::functions::exponential(6, slope_1)); // 7.5 + 1.75^(6 - 4) * 2
-    ASSERT_FLOAT_EQ(18.21875, llmr::functions::exponential(7, slope_1)); // 7.5 + 1.75^(7 - 4) * 2
-    ASSERT_FLOAT_EQ(20, llmr::functions::exponential(8, slope_1)); // 7.5 + 1.75^(8 - 4) * 2  ==> clamped to 20
-    ASSERT_FLOAT_EQ(20, llmr::functions::exponential(20, slope_1)); // 7.5 + 1.75^(20 - 4) * 2 ==> clamped to 20
-
-    // Invalid number of arguments ==> 0
-    std::vector<float> slope_2 = {  };
-    ASSERT_FLOAT_EQ(0, llmr::functions::exponential(3, slope_2));
-    ASSERT_FLOAT_EQ(0, llmr::functions::exponential(8, slope_2));
-    ASSERT_FLOAT_EQ(0, llmr::functions::exponential(10, slope_2));
+    llmr::ExponentialFunction<float> slope_1(/* val */ 7.5, /* z_base */ 4, /* slope */ 2, /* min */ 7.5, /* max */ 20);
+    ASSERT_FLOAT_EQ(8.6428576, slope_1.evaluate(3)); // 7.5 + 1.75^(3 - 4) * 2
+    ASSERT_FLOAT_EQ(9.5, slope_1.evaluate(4)); // 7.5 + 1.75^(4 - 4) * 2
+    ASSERT_FLOAT_EQ(10.145751, slope_1.evaluate(4.5)); // 7.5 + 1.75^(4.5 - 4) * 2
+    ASSERT_FLOAT_EQ(11, slope_1.evaluate(5)); // 7.5 + 1.75^(5 - 4) * 2
+    ASSERT_FLOAT_EQ(13.625, slope_1.evaluate(6)); // 7.5 + 1.75^(6 - 4) * 2
+    ASSERT_FLOAT_EQ(18.21875, slope_1.evaluate(7)); // 7.5 + 1.75^(7 - 4) * 2
+    ASSERT_FLOAT_EQ(20, slope_1.evaluate(8)); // 7.5 + 1.75^(8 - 4) * 2  ==> clamped to 20
+    ASSERT_FLOAT_EQ(20, slope_1.evaluate(20)); // 7.5 + 1.75^(20 - 4) * 2 ==> clamped to 20
 }