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
|
import numpy as np
import matplotlib.pyplot as plt
import sklearn.preprocessing
class Model:
def __init__(self, datafilename="../data.csv", thetafilename="./theta"):
self.datafilename = datafilename
self.thetafilename = thetafilename
self.theta1, self.theta0 = self._read_theta()
self.xs, self.ys = self._read_data()
def train(self, alpha=1, epoch=100):
self._normalize_data()
for _ in range(epoch):
next_theta0 = self.theta0 - alpha * self._partial_theta0()
next_theta1 = self.theta1 - alpha * self._partial_theta1()
self.theta0 = next_theta0
self.theta1 = next_theta1
def write_theta(self):
with open(self.thetafilename, "w") as file:
file.write("{},{}".format(str(self.theta1), str(self.theta0)))
def hypothesis(self, x):
return x * self.theta1 + self.theta0
def cost(self):
return (1 / (2 * len(self.xs))) * sum([(self.hypothesis(x) - y) ** 2
for x, y in zip(self.xs, self.ys)])
def plot(self, plot_data=True, plot_model=True):
self.fig, self.ax = plt.subplots()
if plot_data:
self._plot_data()
if plot_model:
self._plot_model()
plt.show()
def _plot_data(self):
self.ax.scatter(self.xs, self.ys)
def _plot_model(self):
line_xs = [self.xs.min(), self.xs.max()]
line_ys = [self.hypothesis(x) for x in line_xs]
self.ax.plot(line_xs, line_ys)
def _partial_theta1(self):
return sum([(self.hypothesis(x) - y) * x
for x, y in zip(self.xs, self.ys)]) / len(self.xs)
def _partial_theta0(self):
return sum([self.hypothesis(x) - y
for x, y in zip(self.xs, self.ys)]) / len(self.xs)
def _normalize_data(self):
self.xs, self.ys = sklearn.preprocessing.normalize([self.xs, self.ys])
def _read_theta(self):
try:
with open(self.thetafilename, "r") as file:
strs = file.read().strip().split(",")
if len(strs) != 2:
raise "wrong theta file format"
return float(strs[0]), float(strs[1])
except IOError:
print(self.thetafilename, "do not exist")
def _read_data(self):
try:
data = np.genfromtxt(self.datafilename, delimiter=",")[1:]
return data[:, 0], data[:, 1]
except IOError:
print(self.datafilename, "do not exist")
if __name__ == "__main__":
m = Model()
m.train()
m.write_theta()
|
