In [53]:

    

# Необходмые команды импорта.
import sys
sys.path.append('../physlearn/')
sys.path.append('../source')
import numpy as np
from numpy import linalg as LA
import tensorflow as tf
from matplotlib import pylab as plt
import numpy.random as rand
from physlearn.NeuralNet.NeuralNet import NeuralNet
from physlearn.Optimizer.NelderMead.NelderMead import NelderMead
import math_util
import math

from mpl_toolkits.mplot3d.axes3d import Axes3D
%matplotlib notebook


def grid_vect(a, b, N):
    x = np.linspace(a, b, N, endpoint=True)
    h = x[1] - x[0]
    return np.mgrid[a:b+h:h, a:b+h:h].reshape(2,-1).T

def get_meshgrid(a,b,N):
    x = np.linspace(a, b, N, endpoint=True)
    return np.meshgrid(x,x)


a = -2*math.pi
b = 2*math.pi
k = 1
sigmoid_ammount = 35
m_1 = 75
space_dim = 2
iterations = int(5e5)
max_eps = 1e-10

xy = grid_vect(a, b, m_1)
x_np = np.transpose(xy)

m = xy[:,0].size

net = NeuralNet(-5,5)
net.add_input_layer(space_dim)
net.add(sigmoid_ammount, tf.sigmoid)
net.add_output_layer(1, net.linear)
net.compile()
net.set_random_matrixes()
net_out = net.return_graph()
sess = net.return_session()
dim = net.return_unroll_dim()

f = tf.cos(k*tf.sqrt(tf.reduce_sum(tf.square(net.x), axis = 0)))
J = (tf.reduce_sum(tf.square(f - tf.reduce_sum(net_out, axis = 0))))*(1/m) 

#f = tf.cos(k*x_tf)

def COST(params):
    net.roll_matrixes(params)
    res = net.calc(J, {net.x:x_np})
    return res


opt_nm = NelderMead(-2.5,2.5, progress_bar='tqdm')
opt_nm.set_epsilon_and_sd(0.3, 100)

def opt(J, dim, n_it, eps):
    optimisation_result = opt_nm.optimize(J, dim, n_it, eps)
    return optimisation_result


    

In [54]:

    

nn_val = net.run(x_np)
nn_val = nn_val.reshape(1,nn_val.size)
print(nn_val.shape)
fig = plt.figure(figsize=(14,6))
ax = fig.add_subplot(1, 2, 1, projection='3d')
ax.plot_wireframe(xy[:,0], xy[:,1], nn_val)


    

    




(1, 5625)






    















    












    
Out[54]:





<mpl_toolkits.mplot3d.art3d.Line3DCollection at 0x1bdafd23828>

In [55]:

    

optimisation_result = opt(COST, dim, iterations, max_eps)
print("J after optimisation: ", COST(optimisation_result.x))
print("Информация: ", optimisation_result)


    

    




100%|█████████████████████████████████████████████████████████████████████████| 100000/100000 [04:00<00:00, 415.99it/s]






    




J after optimisation:  0.08246693905058346
Информация:  Is converge: False
Amount of iterations: 100000
Total time: 240.39 s
Reached function value: 0.08246693905058346
Reason of break: Maximum iterations reached

In [56]:

    

nn_val = net.run(x_np)
nn_val = nn_val.reshape(1,nn_val.size)
print(nn_val.shape)
fig = plt.figure(figsize=(14,6))
ax = fig.add_subplot(1, 2, 1, projection='3d')
ax.plot_wireframe(xy[:,0], xy[:,1], nn_val)


    

    




(1, 5625)






    















    












    
Out[56]:





<mpl_toolkits.mplot3d.art3d.Line3DCollection at 0x1bdbcb19d68>

In [57]:

    

f_true = net.calc(f, {net.x:x_np})
f_true = f_true.reshape(1,f_true.size)
print(f_true.shape)
fig = plt.figure(figsize=(14,6))
ax = fig.add_subplot(1, 2, 1, projection='3d')
ax.plot_wireframe(xy[:,0], xy[:,1], f_true)


    

    




(1, 5625)






    















    












    
Out[57]:





<mpl_toolkits.mplot3d.art3d.Line3DCollection at 0x1bdb5543b70>

In [58]:

    

error = nn_val - f_true
print(error.shape)
fig = plt.figure(figsize=(14,6))
ax = fig.add_subplot(1, 2, 1, projection='3d')
ax.plot_wireframe(xy[:,0], xy[:,1], error)


    

    




(1, 5625)






    















    












    
Out[58]:





<mpl_toolkits.mplot3d.art3d.Line3DCollection at 0x1bdb56526a0>

In [59]:

    

J_fin = optimisation_result.cost_function
mse = math_util.MSE(error)
std_err = math_util.std_err(error)
print(J_fin, ' ', mse, ' ', std_err)


    

    




0.08246693905058346   0.0824669390505835   0.2871961045114888

In [60]:

    

COST(optimisation_result.x)


    

    
Out[60]:





0.08246693905058346