From b95084dd4d615c3a685eebca83df4c16e7133910 Mon Sep 17 00:00:00 2001
From: Tristan Murphy <72839119+inflectrix@users.noreply.github.com>
Date: Thu, 18 Apr 2024 14:53:59 +0000
Subject: [PATCH 1/2] create custom activations example

---
 examples/custom_activation.rs | 92 +++++++++++++++++++++++++++++++++++
 src/topology/activation.rs    |  4 +-
 2 files changed, 94 insertions(+), 2 deletions(-)
 create mode 100644 examples/custom_activation.rs

diff --git a/examples/custom_activation.rs b/examples/custom_activation.rs
new file mode 100644
index 0000000..f52882b
--- /dev/null
+++ b/examples/custom_activation.rs
@@ -0,0 +1,92 @@
+//! An example implementation of a custom activation function.
+
+use neat::*;
+use rand::prelude::*;
+
+#[derive(DivisionReproduction, RandomlyMutable, Clone)]
+struct AgentDNA {
+    network: NeuralNetworkTopology<2, 2>,
+}
+
+impl Prunable for AgentDNA {}
+
+impl GenerateRandom for AgentDNA {
+    fn gen_random(rng: &mut impl Rng) -> Self {
+        Self {
+            network: NeuralNetworkTopology::new(0.01, 3, rng),
+        }
+    }
+}
+
+fn fitness(g: &AgentDNA) -> f32 {
+    let network: NeuralNetwork<2, 2> = NeuralNetwork::from(&g.network);
+    let mut fitness = 0.;
+    let mut rng = rand::thread_rng();
+
+    for _ in 0..50 {
+        let n = rng.gen::<f32>();
+        let n2 = rng.gen::<f32>();
+
+        let expected = if (n + n2) / 2. >= 0.5 {
+            0
+        } else {
+            1
+        };
+
+        let result = network.predict([n, n2]);
+        network.flush_state();
+
+        // partial_cmp chance of returning None in this smh
+        let result = result.iter().max_index();
+
+        if result == expected {
+            fitness += 1.;
+        } else {
+            fitness -= 1.;
+        }
+    }
+
+    fitness
+}
+
+#[cfg(feature = "serde")]
+fn serde_nextgen(rewards: Vec<(AgentDNA, f32)>) -> Vec<AgentDNA> {
+    let max = rewards
+        .iter()
+        .max_by(|(_, ra), (_, rb)| ra.total_cmp(rb))
+        .unwrap();
+
+    let ser = NNTSerde::from(&max.0.network);
+    let data = serde_json::to_string_pretty(&ser).unwrap();
+    std::fs::write("best-agent.json", data).expect("Failed to write to file");
+
+    division_pruning_nextgen(rewards)
+}
+
+fn main() {
+    let log_activation = activation_fn!(f32::log10);
+    register_activation(log_activation);
+
+    #[cfg(not(feature = "rayon"))]
+    let mut rng = rand::thread_rng();
+
+    let mut sim = GeneticSim::new(
+        #[cfg(not(feature = "rayon"))]
+        Vec::gen_random(&mut rng, 100),
+
+        #[cfg(feature = "rayon")]
+        Vec::gen_random(100),
+
+        fitness,
+
+        #[cfg(not(feature = "serde"))]
+        division_pruning_nextgen,
+
+        #[cfg(feature = "serde")]
+        serde_nextgen,
+    );
+
+    for _ in 0..200 {
+        sim.next_generation();
+    }
+}
\ No newline at end of file
diff --git a/src/topology/activation.rs b/src/topology/activation.rs
index a711851..5bf9540 100644
--- a/src/topology/activation.rs
+++ b/src/topology/activation.rs
@@ -15,11 +15,11 @@ use crate::NeuronLocation;
 #[macro_export]
 macro_rules! activation_fn {
     ($F: path) => {
-        ActivationFn::new(Arc::new($F), ActivationScope::default(), stringify!($F).into())
+        ActivationFn::new(std::sync::Arc::new($F), ActivationScope::default(), stringify!($F).into())
     };
 
     ($F: path, $S: expr) => {
-        ActivationFn::new(Arc::new($F), $S, stringify!($F).into())
+        ActivationFn::new(std::sync::Arc::new($F), $S, stringify!($F).into())
     };
 
     {$($F: path),*} => {

From 7c31f30f88bcd0221e628daa8b0d0530f5c46523 Mon Sep 17 00:00:00 2001
From: Tristan Murphy <tristan.murphy@outlook.com>
Date: Thu, 13 Jun 2024 10:54:22 -0400
Subject: [PATCH 2/2] cargo fmt

---
 examples/custom_activation.rs | 12 ++----------
 1 file changed, 2 insertions(+), 10 deletions(-)

diff --git a/examples/custom_activation.rs b/examples/custom_activation.rs
index f52882b..bc6aae2 100644
--- a/examples/custom_activation.rs
+++ b/examples/custom_activation.rs
@@ -27,11 +27,7 @@ fn fitness(g: &AgentDNA) -> f32 {
         let n = rng.gen::<f32>();
         let n2 = rng.gen::<f32>();
 
-        let expected = if (n + n2) / 2. >= 0.5 {
-            0
-        } else {
-            1
-        };
+        let expected = if (n + n2) / 2. >= 0.5 { 0 } else { 1 };
 
         let result = network.predict([n, n2]);
         network.flush_state();
@@ -73,15 +69,11 @@ fn main() {
     let mut sim = GeneticSim::new(
         #[cfg(not(feature = "rayon"))]
         Vec::gen_random(&mut rng, 100),
-
         #[cfg(feature = "rayon")]
         Vec::gen_random(100),
-
         fitness,
-
         #[cfg(not(feature = "serde"))]
         division_pruning_nextgen,
-
         #[cfg(feature = "serde")]
         serde_nextgen,
     );
@@ -89,4 +81,4 @@ fn main() {
     for _ in 0..200 {
         sim.next_generation();
     }
-}
\ No newline at end of file
+}