1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146


extern crate rand;
extern crate uuid;

use std::collections::HashMap;
use rand::Rng;
use uuid::Uuid;

use super::activation::Activation;
use super::neurontype::NeuronType;



/// Neuron is a wrapper around a neuron providing only what is needed for a neuron to be added 
/// to the NEAT graph, while the neuron encapsulates the neural network logic for the specific nodetype,
/// Some neurons like an LSTM require more variables and different interal activation logic, 
/// so encapsulating that within a normal node on the graph would be misplaced.
#[derive(Deserialize, Serialize, Debug)]
pub struct Neuron {
    pub innov: Uuid,
    pub outgoing: Vec<Uuid>,
    pub incoming: HashMap<Uuid, Option<f32>>,
    pub bias: f32,
    pub value: f32,
    pub d_value: f32,
    pub state: f32,
    pub error: f32,
    pub activation: Activation,
    pub neuron_type: NeuronType
}



impl Neuron {


    pub fn new(innov: Uuid, neuron_type: NeuronType, activation: Activation) -> Self {
        Neuron {
            innov,
            outgoing: Vec::new(),
            incoming: HashMap::new(),
            bias: rand::thread_rng().gen::<f32>(),
            value: 0.0,
            d_value: 0.0,
            state: 0.0,
            error: 0.0,
            activation,
            neuron_type,
        }
    }



    /// Return this struct as a raw mutable pointer - consumes the struct
    pub fn as_mut_ptr(self) -> *mut Neuron {
        Box::into_raw(Box::new(self))
    }



    /// figure out if this node can be calculated, meaning all of the 
    /// nodes pointing to it have given this node their output values.
    /// If they have, this node is ready to be activated
    #[inline]
    pub fn is_ready(&mut self) -> bool {
        self.incoming
            .values()
            .all(|x| x.is_some())
    }


    
    /// 𝜎(Σ(w * i) + b)
    /// activate this node by calling the underlying neuron's logic for activation
    /// given the hashmap of <incoming edge innov, Option<incoming Neuron output value>>
    #[inline]
    pub fn activate(&mut self) {
        self.state = self.incoming
            .values()
            .fold(self.bias, |sum, curr| {
                match curr {
                    Some(x) => sum + x,
                    None => panic!("Cannot activate node.")
                }
            });
        if self.activation != Activation::Softmax {
            self.value = self.activation.activate(self.state);
            self.d_value = self.activation.deactivate(self.state);
        }
    }



    /// deactivate this node by calling the underlying neuron's logic to compute
    /// the gradient of the original output value 
    #[inline]
    pub fn deactivate(&self) -> f32 {
        self.activation.deactivate(self.state)
    }



    /// each Neuron has a base layer of reset which needs to happen 
    /// but on top of that each neuron might need to do more interanally
    #[inline]
    pub fn reset_neuron(&mut self) {
        self.error = 0.0;
        for (_, val) in self.incoming.iter_mut() {
            *val = None;
        }
    }


}



impl Clone for Neuron {
    fn clone(&self) -> Self { 
        Neuron {
            innov: self.innov,
            outgoing: self.outgoing
                .iter()
                .map(|x| *x)
                .collect(),
            incoming: self.incoming
                .iter()
                .map(|(key, _)| (*key, None))
                .collect(),
            state: 0.0,
            value: 0.0,
            d_value: 0.0,
            error: 0.0,
            bias: self.bias.clone(),
            activation: self.activation.clone(),
            neuron_type: self.neuron_type.clone(),
        }
    }
}


impl PartialEq for Neuron {
    fn eq(&self, other: &Self) -> bool {
        self.innov == other.innov
    }
}