Make_Results.py

'''
Copyright 2020 Amanpreet Singh,
               Martin Bauer,
               Sarang Joshi

Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:

1. Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.

2. Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

'''



from Geometry import *
from Distributions import *
from ICNN import *
from Draw_Graphs_Supp import *
from Datasets import *
import matplotlib.pyplot as plt

# Specify the GPU to use.
os.environ["CUDA_VISIBLE_DEVICES"] = "0"


dim = 2

# PARAMETERS for the network
hidden_layers = 5
npl = 128
epochs = 10000

# Regularisation constant
k = 0

# Starting from this epoch till the end with step size same as this variable
save_after_e = 10000

# Flags as used in the training of network.
Period_Boundary_Flag = False
deformation_flag = False

# Functions g and f. Note in discrete formulation we don't have any function f. function g is a unit gaussian.
func_g = gauss

# train_data = FunnyDistDataset('/hdscratch/Monge_Ampere/FunnyDist/oit-random-master/funnydist_samples.txt')
train_data = MultipleGaussianMixtureDataset(4, 10000, [[-0.5, -0.5], [0.5, 0.5], [0.5, -0.5], [-0.5, 0.5]],
                                            [[0.25**2, 0.25**2], [0.25**2, 0.25**2], [0.25**2, 0.25**2],
                                             [0.25**2, 0.25**2]])
data = train_data[:][0]


# mu = data[:, 0:dim].mean(axis=0)
# cov = np.cov(data[:, 0:dim].T)


# def func_g(x, c):
#     sum = torch.exp(gauss_md_2ms(x))
#     return sum

# func_g = gauss_md_2ms

# FILE PATHS
# FILE PATHS
model_path = './Models/4c/'.format(dim)
graph_path = './Graphs/4c/'.format(dim)
model_name = 'model_hl_' + str(hidden_layers) + '_npl_' + str(npl) + '_k_' + str(k) + '/'

model_path = model_path + model_name
graph_path = graph_path + model_name

os.makedirs(graph_path, exist_ok=True)


fig = plt.figure(frameon=False)
ax = plt.Axes(fig, [0, 0, 1, 1])
ax.set_axis_off()
fig.add_axes(ax)
ax.scatter(data[:, 0], data[:, 1], c='k', s=0.1)
ax.set_xlim(-2, 2)
ax.set_ylim(-2, 2)
fig.savefig(graph_path + 'Training_data.png')
plt.close(fig)


# Define the model and put that on the GPU
model = ICNN(2, 1, hidden_layers, npl)
model = model.cuda()

# Call the function to generate graphs
make_graphs_all_epochs(graph_path, model_path, model, save_after_e, epochs, func_g,
                       norm_const=None, periodic_boundary_flag=Period_Boundary_Flag, deformation_flag=deformation_flag)

# For debugging
print("All Done")