-
Notifications
You must be signed in to change notification settings - Fork 14
/
analyze-effect-tau.py
197 lines (175 loc) · 9.88 KB
/
analyze-effect-tau.py
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
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
# Symplectic ODE-Net | 2019
# Yaofeng Desmond Zhong, Biswadip Dey, Amit Chakraborty
# code structure follows the style of HNN by Sam Greydanus
# https://github.com/greydanus/hamiltonian-nn
# This file is a script version of 'analyze-effect-tau.ipynb'
# Cells are seperated by the vscode convention '#%%'
#%%
import torch, time, sys
import numpy as np
import matplotlib.pyplot as plt
import scipy.integrate
solve_ivp = scipy.integrate.solve_ivp
EXPERIMENT_DIR = './experiment-single-embed/'
sys.path.append(EXPERIMENT_DIR)
from data import get_dataset, arrange_data, get_field
from nn_models import MLP, PSD
from symoden import SymODEN_T
from utils import L2_loss, from_pickle
import imageio
#%%
DPI = 600
FORMAT = 'png'
def get_args():
return {'num_angle': 1,
'nonlinearity': 'tanh',
'name': 'pend',
'seed': 0,
'save_dir': './{}'.format(EXPERIMENT_DIR),
'fig_dir': './figures',
'num_points': 5,
'gpu': 0,
'solver': 'dopri5'}
class ObjectView(object):
def __init__(self, d): self.__dict__ = d
args = ObjectView(get_args())
#%%
device = torch.device('cuda:' + str(args.gpu) if torch.cuda.is_available() else 'cpu')
def get_model(args, baseline, structure, naive, num_points, solver):
M_net = PSD(2*args.num_angle, 300, args.num_angle).to(device)
g_net = MLP(2*args.num_angle, 200, args.num_angle).to(device)
if structure == False:
if naive and baseline:
raise RuntimeError('argument *baseline* and *naive* cannot both be true')
elif naive:
input_dim = 4 * args.num_angle
output_dim = 3 * args.num_angle
nn_model = MLP(input_dim, 800, output_dim, args.nonlinearity).to(device)
model = SymODEN_T(args.num_angle, H_net=nn_model, device=device, baseline=baseline, naive=naive)
elif baseline:
input_dim = 4 * args.num_angle
output_dim = 2 * args.num_angle
nn_model = MLP(input_dim, 600, output_dim, args.nonlinearity).to(device)
model = SymODEN_T(args.num_angle, H_net=nn_model, M_net=M_net, device=device, baseline=baseline, naive=naive)
else:
input_dim = 3 * args.num_angle
output_dim = 1
nn_model = MLP(input_dim, 500, output_dim, args.nonlinearity).to(device)
model = SymODEN_T(args.num_angle, H_net=nn_model, M_net=M_net, g_net=g_net, device=device, baseline=baseline, naive=naive)
elif structure == True and baseline ==False and naive==False:
V_net = MLP(2*args.num_angle, 50, 1).to(device)
model = SymODEN_T(args.num_angle, M_net=M_net, V_net=V_net, g_net=g_net, device=device, baseline=baseline, structure=True).to(device)
else:
raise RuntimeError('argument *structure* is set to true, no *baseline* or *naive*!')
if naive:
label = '-naive_ode'
elif baseline:
label = '-baseline_ode'
else:
label = '-hnn_ode'
struct = '-struct' if structure else ''
path = '{}/{}{}{}-{}-p{}.tar'.format(args.save_dir, args.name, label, struct, solver, num_points)
model.load_state_dict(torch.load(path, map_location=device))
path = '{}/{}{}{}-{}-p{}-stats.pkl'.format(args.save_dir, args.name, label, struct, solver, num_points)
stats = from_pickle(path)
return model, stats
model_rk4_p2, stats_rk4_p2 = get_model(args, baseline=False, structure=True, naive=False, num_points=2, solver='rk4')
model_rk4_p3, stats_rk4_p3 = get_model(args, baseline=False, structure=True, naive=False, num_points=3, solver='rk4')
model_rk4_p4, stats_rk4_p4 = get_model(args, baseline=False, structure=True, naive=False, num_points=4, solver='rk4')
model_rk4_p5, stats_rk4_p5 = get_model(args, baseline=False, structure=True, naive=False, num_points=5, solver='rk4')
model_rk4_p6, stats_rk4_p6 = get_model(args, baseline=False, structure=True, naive=False, num_points=6, solver='rk4')
model_dopri5_p2, stats_dopri5_p2 = get_model(args, baseline=False, structure=True, naive=False, num_points=2, solver='dopri5')
model_dopri5_p3, stats_dopri5_p3 = get_model(args, baseline=False, structure=True, naive=False, num_points=3, solver='dopri5')
model_dopri5_p4, stats_dopri5_p4 = get_model(args, baseline=False, structure=True, naive=False, num_points=4, solver='dopri5')
model_dopri5_p5, stats_dopri5_p5 = get_model(args, baseline=False, structure=True, naive=False, num_points=5, solver='dopri5')
model_dopri5_p6, stats_dopri5_p6 = get_model(args, baseline=False, structure=True, naive=False, num_points=6, solver='dopri5')
print('stats_rk4_p2')
print('Final trajectory train loss {:.4e} +/- {:.4e}\nFinal trajectory test loss {:.4e} +/- {:.4e}'
.format(np.mean(stats_rk4_p2['traj_train_loss']), np.std(stats_rk4_p2['traj_train_loss']),
np.mean(stats_rk4_p2['traj_test_loss']), np.std(stats_rk4_p2['traj_test_loss'])))
print('stats_rk4_p3')
print('Final trajectory train loss {:.4e} +/- {:.4e}\nFinal trajectory test loss {:.4e} +/- {:.4e}'
.format(np.mean(stats_rk4_p3['traj_train_loss']), np.std(stats_rk4_p3['traj_train_loss']),
np.mean(stats_rk4_p3['traj_test_loss']), np.std(stats_rk4_p3['traj_test_loss'])))
print('stats_rk4_p4')
print('Final trajectory train loss {:.4e} +/- {:.4e}\nFinal trajectory test loss {:.4e} +/- {:.4e}'
.format(np.mean(stats_rk4_p4['traj_train_loss']), np.std(stats_rk4_p4['traj_train_loss']),
np.mean(stats_rk4_p4['traj_test_loss']), np.std(stats_rk4_p4['traj_test_loss'])))
print('stats_rk4_p5')
print('Final trajectory train loss {:.4e} +/- {:.4e}\nFinal trajectory test loss {:.4e} +/- {:.4e}'
.format(np.mean(stats_rk4_p5['traj_train_loss']), np.std(stats_rk4_p5['traj_train_loss']),
np.mean(stats_rk4_p5['traj_test_loss']), np.std(stats_rk4_p5['traj_test_loss'])))
print('stats_rk4_p6')
print('Final trajectory train loss {:.4e} +/- {:.4e}\nFinal trajectory test loss {:.4e} +/- {:.4e}'
.format(np.mean(stats_rk4_p6['traj_train_loss']), np.std(stats_rk4_p6['traj_train_loss']),
np.mean(stats_rk4_p6['traj_test_loss']), np.std(stats_rk4_p6['traj_test_loss'])))
print('')
print('stats_dopri5_p2')
print('Final trajectory train loss {:.4e} +/- {:.4e}\nFinal trajectory test loss {:.4e} +/- {:.4e}'
.format(np.mean(stats_dopri5_p2['traj_train_loss']), np.std(stats_dopri5_p2['traj_train_loss']),
np.mean(stats_dopri5_p2['traj_test_loss']), np.std(stats_dopri5_p2['traj_test_loss'])))
print('stats_dopri5_p3')
print('Final trajectory train loss {:.4e} +/- {:.4e}\nFinal trajectory test loss {:.4e} +/- {:.4e}'
.format(np.mean(stats_dopri5_p3['traj_train_loss']), np.std(stats_dopri5_p3['traj_train_loss']),
np.mean(stats_dopri5_p3['traj_test_loss']), np.std(stats_dopri5_p3['traj_test_loss'])))
print('stats_dopri5_p4')
print('Final trajectory train loss {:.4e} +/- {:.4e}\nFinal trajectory test loss {:.4e} +/- {:.4e}'
.format(np.mean(stats_dopri5_p4['traj_train_loss']), np.std(stats_dopri5_p4['traj_train_loss']),
np.mean(stats_dopri5_p4['traj_test_loss']), np.std(stats_dopri5_p4['traj_test_loss'])))
print('stats_dopri5_p5')
print('Final trajectory train loss {:.4e} +/- {:.4e}\nFinal trajectory test loss {:.4e} +/- {:.4e}'
.format(np.mean(stats_dopri5_p5['traj_train_loss']), np.std(stats_dopri5_p5['traj_train_loss']),
np.mean(stats_dopri5_p5['traj_test_loss']), np.std(stats_dopri5_p5['traj_test_loss'])))
print('stats_dopri5_p6')
print('Final trajectory train loss {:.4e} +/- {:.4e}\nFinal trajectory test loss {:.4e} +/- {:.4e}'
.format(np.mean(stats_dopri5_p6['traj_train_loss']), np.std(stats_dopri5_p6['traj_train_loss']),
np.mean(stats_dopri5_p6['traj_test_loss']), np.std(stats_dopri5_p6['traj_test_loss'])))
#%%
#%%
us = [0.0]
data = get_dataset(seed=args.seed, timesteps=40,
save_dir=args.save_dir, us=us, samples=128)
pred_x, pred_t_eval = data['x'], data['t']
from torchdiffeq import odeint
def get_pred_loss(pred_x, pred_t_eval, model):
pred_x = torch.tensor(pred_x, requires_grad=True, dtype=torch.float32).to(device)
pred_t_eval = torch.tensor(pred_t_eval, requires_grad=True, dtype=torch.float32).to(device)
pred_loss = []
for i in range(pred_x.shape[0]):
pred_x_hat = odeint(model, pred_x[i, 0, :, :], pred_t_eval, method='rk4')
pred_loss.append((pred_x[i,:,:,:] - pred_x_hat)**2)
pred_loss = torch.cat(pred_loss, dim=1)
pred_loss_per_traj = torch.sum(pred_loss, dim=(0, 2))
return pred_loss_per_traj.detach().cpu().numpy()
rk4_p2_pred_loss = get_pred_loss(pred_x, pred_t_eval, model_rk4_p2)
rk4_p3_pred_loss = get_pred_loss(pred_x, pred_t_eval, model_rk4_p3)
rk4_p4_pred_loss = get_pred_loss(pred_x, pred_t_eval, model_rk4_p4)
rk4_p5_pred_loss = get_pred_loss(pred_x, pred_t_eval, model_rk4_p5)
rk4_p6_pred_loss = get_pred_loss(pred_x, pred_t_eval, model_rk4_p6)
dopri5_p2_pred_loss = get_pred_loss(pred_x, pred_t_eval, model_dopri5_p2)
dopri5_p3_pred_loss = get_pred_loss(pred_x, pred_t_eval, model_dopri5_p3)
dopri5_p4_pred_loss = get_pred_loss(pred_x, pred_t_eval, model_dopri5_p4)
dopri5_p5_pred_loss = get_pred_loss(pred_x, pred_t_eval, model_dopri5_p5)
dopri5_p6_pred_loss = get_pred_loss(pred_x, pred_t_eval, model_dopri5_p6)
#%%
print('stats_rk4_p2')
print('Prediction loss {:.4e}'.format(np.mean(rk4_p2_pred_loss)))
print('stats_rk4_p3')
print('Prediction loss {:.4e}'.format(np.mean(rk4_p3_pred_loss)))
print('stats_rk4_p4')
print('Prediction loss {:.4e}'.format(np.mean(rk4_p4_pred_loss)))
print('stats_rk4_p5')
print('Prediction loss {:.4e}'.format(np.mean(rk4_p5_pred_loss)))
print('stats_rk4_p6')
print('Prediction loss {:.4e}'.format(np.mean(rk4_p6_pred_loss)))
print('stats_dopri5_p2')
print('Prediction loss {:.4e}'.format(np.mean(dopri5_p2_pred_loss)))
print('stats_dopri5_p3')
print('Prediction loss {:.4e}'.format(np.mean(dopri5_p3_pred_loss)))
print('stats_dopri5_p4')
print('Prediction loss {:.4e}'.format(np.mean(dopri5_p4_pred_loss)))
print('stats_dopri5_p5')
print('Prediction loss {:.4e}'.format(np.mean(dopri5_p5_pred_loss)))
print('stats_dopri5_p6')
print('Prediction loss {:.4e}'.format(np.mean(dopri5_p6_pred_loss)))
# %%