lowrank_3d.py

import torch
import numpy as np
import torch.nn as nn
import torch.nn.functional as F

import matplotlib.pyplot as plt
from utilities3 import *

import operator
from functools import reduce
from functools import partial

from timeit import default_timer
import scipy.io

torch.manual_seed(0)
np.random.seed(0)


################################################################
# 3d lowrank layers
################################################################

class LowRank2d(nn.Module):
    def __init__(self, in_channels, out_channels, n, ker_width, rank):
        super(LowRank2d, self).__init__()
        self.in_channels = in_channels
        self.out_channels = out_channels
        self.n = n
        self.rank = rank

        self.phi = DenseNet([in_channels, ker_width, in_channels*out_channels*rank], torch.nn.ReLU)
        self.psi = DenseNet([in_channels, ker_width, in_channels*out_channels*rank], torch.nn.ReLU)


    def forward(self, v):
        batch_size = v.shape[0]

        phi_eval = self.phi(v).reshape(batch_size, self.n, self.out_channels, self.in_channels, self.rank)
        psi_eval = self.psi(v).reshape(batch_size, self.n, self.out_channels, self.in_channels, self.rank)

        # print(psi_eval.shape, v.shape, phi_eval.shape)
        v = torch.einsum('bnoir,bni,bmoir->bmo',psi_eval, v, phi_eval)

        return v



class MyNet(torch.nn.Module):
    def __init__(self, n, width=16, ker_width=256, rank=16):
        super(MyNet, self).__init__()
        self.n = n
        self.width = width
        self.ker_width = ker_width
        self.rank = rank

        self.fc0 = nn.Linear(13, self.width)

        self.conv0 = LowRank2d(width, width, n, ker_width, rank)
        self.conv1 = LowRank2d(width, width, n, ker_width, rank)
        self.conv2 = LowRank2d(width, width, n, ker_width, rank)
        self.conv3 = LowRank2d(width, width, n, ker_width, rank)

        self.w0 = nn.Linear(self.width, self.width)
        self.w1 = nn.Linear(self.width, self.width)
        self.w2 = nn.Linear(self.width, self.width)
        self.w3 = nn.Linear(self.width, self.width)
        self.bn0 = torch.nn.BatchNorm1d(self.width)
        self.bn1 = torch.nn.BatchNorm1d(self.width)
        self.bn2 = torch.nn.BatchNorm1d(self.width)
        self.bn3 = torch.nn.BatchNorm1d(self.width)

        self.fc1 = nn.Linear(self.width, 128)
        self.fc2 = nn.Linear(128, 1)


    def forward(self, x):
        batch_size = x.shape[0]
        size_x, size_y, size_z = x.shape[1], x.shape[2], x.shape[3]
        x = x.view(batch_size, size_x*size_y*size_z, -1)

        x = self.fc0(x)

        x1 = self.conv0(x)
        x2 = self.w0(x)
        x = x1 + x2
        x = self.bn0(x.reshape(-1, self.width)).view(batch_size, size_x*size_y*size_z, self.width)
        x = F.relu(x)
        x1 = self.conv1(x)
        x2 = self.w1(x)
        x = x1 + x2
        x = self.bn1(x.reshape(-1, self.width)).view(batch_size, size_x*size_y*size_z, self.width)
        x = F.relu(x)
        x1 = self.conv2(x)
        x2 = self.w2(x)
        x = x1 + x2
        x = self.bn2(x.reshape(-1, self.width)).view(batch_size, size_x*size_y*size_z, self.width)
        x = F.relu(x)
        x1 = self.conv3(x)
        x2 = self.w3(x)
        x = x1 + x2
        x = self.bn3(x.reshape(-1, self.width)).view(batch_size, size_x*size_y*size_z, self.width)

        x = self.fc1(x)
        x = F.relu(x)
        x = self.fc2(x)
        x = x.view(batch_size, size_x, size_y, size_z)
        return x

class Net2d(nn.Module):
    def __init__(self, width=8, ker_width=128, rank=4):
        super(Net2d, self).__init__()

        self.conv1 = MyNet(n=64*64*40, width=width, ker_width=ker_width, rank=rank)


    def forward(self, x):
        x = self.conv1(x)
        return x


    def count_params(self):
        c = 0
        for p in self.parameters():
            c += reduce(operator.mul, list(p.size()))

        return c

################################################################
# configs
################################################################
# TRAIN_PATH = 'data/ns_data_V10000_N1200_T20.mat'
# TEST_PATH = 'data/ns_data_V10000_N1200_T20.mat'
# TRAIN_PATH = 'data/ns_data_V1000_N1000_train.mat'
# TEST_PATH = 'data/ns_data_V1000_N1000_train_2.mat'
# TRAIN_PATH = 'data/ns_data_V1000_N5000.mat'
# TEST_PATH = 'data/ns_data_V1000_N5000.mat'
TRAIN_PATH = 'data/ns_data_V100_N1000_T50_1.mat'
TEST_PATH = 'data/ns_data_V100_N1000_T50_2.mat'

ntrain = 1000
ntest = 200

batch_size = 2
batch_size2 = batch_size

epochs = 500
learning_rate = 0.0025
scheduler_step = 100
scheduler_gamma = 0.5

print(epochs, learning_rate, scheduler_step, scheduler_gamma)

path = 'ns_lowrank_V100_T40_N'+str(ntrain)+'_ep' + str(epochs)
path_model = 'model/'+path
path_train_err = 'results/'+path+'train.txt'
path_test_err = 'results/'+path+'test.txt'
path_image = 'image/'+path

runtime = np.zeros(2, )
t1 = default_timer()


sub = 1
S = 64
T_in = 10
T = 40

################################################################
# load data
################################################################

reader = MatReader(TRAIN_PATH)
train_a = reader.read_field('u')[:ntrain,::sub,::sub,:T_in]
train_u = reader.read_field('u')[:ntrain,::sub,::sub,T_in:T+T_in]

reader = MatReader(TEST_PATH)
test_a = reader.read_field('u')[-ntest:,::sub,::sub,:T_in]
test_u = reader.read_field('u')[-ntest:,::sub,::sub,T_in:T+T_in]

print(train_u.shape)
print(test_u.shape)
assert (S == train_u.shape[-2])
assert (T == train_u.shape[-1])


a_normalizer = UnitGaussianNormalizer(train_a)
train_a = a_normalizer.encode(train_a)
test_a = a_normalizer.encode(test_a)

y_normalizer = UnitGaussianNormalizer(train_u)
train_u = y_normalizer.encode(train_u)

train_a = train_a.reshape(ntrain,S,S,1,T_in).repeat([1,1,1,T,1])
test_a = test_a.reshape(ntest,S,S,1,T_in).repeat([1,1,1,T,1])

gridx = torch.tensor(np.linspace(0, 1, S), dtype=torch.float)
gridx = gridx.reshape(1, S, 1, 1, 1).repeat([1, 1, S, T, 1])
gridy = torch.tensor(np.linspace(0, 1, S), dtype=torch.float)
gridy = gridy.reshape(1, 1, S, 1, 1).repeat([1, S, 1, T, 1])
gridt = torch.tensor(np.linspace(0, 1, T+1)[1:], dtype=torch.float)
gridt = gridt.reshape(1, 1, 1, T, 1).repeat([1, S, S, 1, 1])

train_a = torch.cat((gridx.repeat([ntrain,1,1,1,1]), gridy.repeat([ntrain,1,1,1,1]),
                       gridt.repeat([ntrain,1,1,1,1]), train_a), dim=-1)
test_a = torch.cat((gridx.repeat([ntest,1,1,1,1]), gridy.repeat([ntest,1,1,1,1]),
                       gridt.repeat([ntest,1,1,1,1]), test_a), dim=-1)

train_loader = torch.utils.data.DataLoader(torch.utils.data.TensorDataset(train_a, train_u), batch_size=batch_size, shuffle=True)
test_loader = torch.utils.data.DataLoader(torch.utils.data.TensorDataset(test_a, test_u), batch_size=batch_size, shuffle=False)

t2 = default_timer()

print('preprocessing finished, time used:', t2-t1)
device = torch.device('cuda')


################################################################
# training and evaluation
################################################################
model = Net2d().cuda()
# model = torch.load('model/ns_fourier_V100_N1000_ep100_m8_w20')

print(model.count_params())
optimizer = torch.optim.Adam(model.parameters(), lr=learning_rate, weight_decay=1e-4)
scheduler = torch.optim.lr_scheduler.StepLR(optimizer, step_size=scheduler_step, gamma=scheduler_gamma)


myloss = LpLoss(size_average=False)
y_normalizer.cuda()
for ep in range(epochs):
    model.train()
    t1 = default_timer()
    train_mse = 0
    train_l2 = 0
    for x, y in train_loader:
        x, y = x.cuda(), y.cuda()

        optimizer.zero_grad()
        out = model(x)

        mse = F.mse_loss(out, y, reduction='mean')
        # mse.backward()

        y = y_normalizer.decode(y)
        out = y_normalizer.decode(out)
        l2 = myloss(out.view(batch_size, -1), y.view(batch_size, -1))
        l2.backward()

        optimizer.step()
        train_mse += mse.item()
        train_l2 += l2.item()

    scheduler.step()

    model.eval()
    test_l2 = 0.0
    with torch.no_grad():
        for x, y in test_loader:
            x, y = x.cuda(), y.cuda()

            out = model(x)
            out = y_normalizer.decode(out)
            test_l2 += myloss(out.view(batch_size, -1), y.view(batch_size, -1)).item()

    train_mse /= len(train_loader)
    train_l2 /= ntrain
    test_l2 /= ntest

    t2 = default_timer()
    print(ep, t2-t1, train_mse, train_l2, test_l2)
# torch.save(model, path_model)


pred = torch.zeros(test_u.shape)
index = 0
test_loader = torch.utils.data.DataLoader(torch.utils.data.TensorDataset(test_a, test_u), batch_size=1, shuffle=False)
with torch.no_grad():
    for x, y in test_loader:
        test_l2 = 0;
        x, y = x.cuda(), y.cuda()

        out = model(x)
        out = y_normalizer.decode(out)
        pred[index] = out

        test_l2 += myloss(out.view(1, -1), y.view(1, -1)).item()
        print(index, test_l2)
        index = index + 1

# scipy.io.savemat('pred/'+path+'.mat', mdict={'pred': pred.cpu().numpy()})