Python get_dot_graph示例

示例#1

文件： test_utils.py 项目： Hagihara-A/dl-scratch-3

    def test__dot_var(self):
        a = Variable(np.array(1.0))
        b = Variable(np.array(1.0))
        c = a + b
        d = square(c)

        actual = get_dot_graph(d)
        self.assertIn(f"{id(d.creator)} -> {id(d)}", actual)
        self.assertIn(f"{id(c)} -> {id(d.creator)}", actual)

示例#2

文件： test_second_order.py 项目： centertong/dezero

    def test_tanh(self):
        x = Variable(np.array(1.0))
        y = F.tanh(x)
        x.name = 'x'
        y.name = 'y'
        y.backward(create_graph=True)

        iters = 0

        for i in range(iters):
            gx = x.grad
            x.cleargrad()
            gx.backward(create_graph=True)

        gx = x.grad
        gx.name = 'gx' + str(iters + 1)
        txt = get_dot_graph(gx)
        with open('test.dot', 'w') as f:
            f.write(txt)

示例#3

文件： models.py 项目： khiro112/dezero

 def plot(self, *inputs, to_file='model.png'):
     y = self.forward(*inputs)
     return utils.get_dot_graph(y, verbose=True, to_file=to_file)

示例#4

'''
計算グラフの可視化は、次のコマンドで行えます
$ python step22.py | dot -T png -o sample.png
'''

import numpy as np
from dezero import Variable
from dezero.utils import get_dot_graph


def goldstein(x0, x1):
    y = (1 + (x0 + x1 + 1)**2 * (19 - 14*x0 + 3*x0**2 - 14*x1 + 6*x0*x1 + 3*x1**2)) *\
        (30 + (2*x0 - 3*x1)**2 * (18 - 32*x0 + 12*x0**2 + 48*x1 - 36*x0*x1 + 27*x1**2))
    return y


x0 = Variable(np.array(1.0))
x1 = Variable(np.array(1.0))
y = goldstein(x0, x1)
y.backward()

x0.name = 'x0'
x1.name = 'x1'
y.name = 'y'
print(get_dot_graph(y))

示例#5

文件： step27.py 项目： S-YOU/deep-learning-from-scratch-3

def my_sin(x, iters=5):
    y = 0
    for i in range(iters):
        t = (-1)**i * x**(2 * i + 1) / math.factorial(2 * i + 1)
        y = y + t
    return y


def my_sin(x, threshould=0.0001):
    y = 0
    for i in range(100000):
        t = (-1)**i * x**(2 * i + 1) / math.factorial(2 * i + 1)
        y = y + t
        if abs(t.data) < threshould:
            break
    return y


x = Variable(np.array(np.pi / 4))
y = my_sin(x)  # , threshould=1e-150)
y.backward()
print('--- approximate sin ---')
print(y.data)
print(x.grad)  # 0.7071032148228457

# 可視化 (dotファイルに保存)
x.name = 'x'
y.name = 'y'
dot = get_dot_graph(y)
with open('my_sin.dot', 'w') as o:
    o.write(dot)

示例#6

文件： tanh.py 项目： leynard007/deep-learning-from-scratch-3

'''
y=tanh(x)のn階微分
計算グラフの可視化は、次のコマンドで行えます
$ python tanh.py | dot -T png -o sample.png
'''
import numpy as np
from dezero import Variable
from dezero.utils import get_dot_graph
import dezero.functions as F

x = Variable(np.array(1.0))
y = F.tanh(x)
x.name = 'x'
y.name = 'y'
y.backward(create_graph=True)

iters = 3

for i in range(iters):
    gx = x.grad
    x.cleargrad()
    gx.backward(create_graph=True)

gx = x.grad
gx.name = 'gx' + str(iters + 1)
print(get_dot_graph(gx))