/
node.py
55 lines (44 loc) · 1.34 KB
/
node.py
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
from base import train, predict
import random
class Node:
def __init__(self, entries, activation_function=None, bias=0):
assert type(entries) == int
assert entries > 0
self.goal_pred = 0
self.weights = [random.triangular(-1, 1)]*entries
self.inputs = [None]*entries
self.pred = 0
self.observers = []
self.activation_function = activation_function
self.bias = bias
def set_weight(self, k, value):
assert k < len(self.weights)
self.weights[k] = value
def check_activation(self):
for i in self.inputs:
if i is None:
break
else:
self.activate()
def transmit(self, k, feature):
assert k < len(self.weights)
self.inputs[k] = feature
self.check_activation()
def transmit_all(self, features):
assert type(features) == list
for k in range(len(features)):
self.transmit(k, features[k])
def connect(self, neuron, k):
assert k < len(self.weights)
self.observers.append(lambda pred: neuron.transmit(k, pred))
def connect(self, cb):
self.observers.append(cb)
def activate(self):
pred = predict(self.inputs, self.weights)
if self.activation_function:
pred = self.activation_function(pred)+self.bias
self.inputs = [None]*len(self.inputs)
for cb in self.observers:
cb(pred)
def train(self, goal_pred, inputs, times=500, alpha=0.01):
self.weights = train(inputs, self.weights, goal_pred, times, alpha)