-
Notifications
You must be signed in to change notification settings - Fork 0
/
2_test_vq_double.py
146 lines (119 loc) · 4.64 KB
/
2_test_vq_double.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
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
"""
VQ PM. review model
"""
import os
import matplotlib.pyplot as plt
import numpy as np
import tensorflow as tf
import pandas as pd
#from vq_model import VAEModel
from vq_model import VAEModel
from scipy.special import softmax
#============================================================================================================
num_views = 4# @params
view_trains = False
PM_FILE_TRAIN = '012018-062019_raw_primemap_train.csv'
PM_FILE_VALID = '012018-062019_raw_primemap_valid.csv'
CHKPT_DIR = 'vq_mod_8nembed'
VER_DIR = 'ver-41'
# Set hyper-parameters.
VALID_FRAC = 10 # @param (10 means 1/10 is test)
BATCH_SIZE = 64
KWARGS ={}
KWARGS['in_channel']=1
KWARGS['main_channel']=32
KWARGS['num_res_blocks']=2
KWARGS['residual_hiddens']=8
# This value is not that important, usually 64 works.
# This will not change the capacity in the information-bottleneck.
KWARGS['embed_dim']=16
# The higher this value, the higher the capacity in the information bottleneck.
KWARGS['n_embed']=8
KWARGS['decay']=0.99
KWARGS['commitment_cost']=0.25
#==============================================================================================================
dir_path = os.path.dirname(os.path.realpath(__file__))
PM_FILE_TRAIN = os.path.join(dir_path,PM_FILE_TRAIN)
PM_FILE_VALID = os.path.join(dir_path,PM_FILE_VALID)
CHKPT_DIR = os.path.join(dir_path,CHKPT_DIR)
train_data = np.array(pd.read_csv(PM_FILE_TRAIN, sep=',',header=None).values,dtype=np.float32)
valid_data = np.array(pd.read_csv(PM_FILE_VALID, sep=',',header=None).values,dtype=np.float32)
def log_and_normalise(data_mat):#,normalizer):
"""Log data range [-18.42, 0.0] (softmax at end)
transform to [0.0 to 1.0]
"""
data_mat = tf.cast(data_mat,dtype=tf.float32)
far_val = tf.constant(18.43,dtype=tf.float32)
data_mat = tf.math.log(data_mat + 1e-8)
data_mat = data_mat/ far_val +1
data_mat= tf.expand_dims(data_mat,-1)
return data_mat
train_dataset_iterator = iter(
tf.data.Dataset.from_tensor_slices(train_data)
.map(log_and_normalise)
.shuffle(100000)
.repeat(-1) # repeat indefinitely
.batch(BATCH_SIZE))
valid_dataset_iterator = iter(
tf.data.Dataset.from_tensor_slices(valid_data)
.map(log_and_normalise)
.repeat(1) # 1 epoch
.batch(BATCH_SIZE))
def get_images(subset='train'):
if subset == 'train':
return next(train_dataset_iterator)
elif subset == 'valid':
return next(valid_dataset_iterator)
vae_all = VAEModel(**KWARGS)
inx = get_images('train')
vae_all(inx)
checkpoint = tf.train.Checkpoint(module = vae_all)
checkpoint.restore(os.path.join(CHKPT_DIR,VER_DIR))
#===========================================VIEW RECONSTRUCTIONS===================================
def do_x_recon(x):
#x_recon,_,_,_=vae_all(x,is_training=False)
x_recon,_,_=vae_all(x,is_training=False)
return x_recon
def revert_transform(image_batch):
image_batch = np.array(image_batch,dtype=np.float32)
images = np.squeeze(image_batch,-1)
images-=1
images*=18.43
images = softmax(images,axis=-1)
return images
xvals = np.linspace(-8,7,16)
train_originals = get_images('train')
train_reconstructions = do_x_recon(train_originals)
valid_originals = get_images('valid')
valid_reconstructions = do_x_recon(valid_originals)
viewinds = np.arange(0,BATCH_SIZE)
for _ in range(5):
#plt.cla()
np.random.shuffle(viewinds)
f = plt.figure(figsize=(8,8))
for i in range(num_views):
if view_trains:
tempY = revert_transform(train_originals[viewinds[i]])
tempY2 =revert_transform(train_reconstructions[viewinds[i]])
else:
tempY = revert_transform(valid_originals[viewinds[i]])
tempY2 =revert_transform(valid_reconstructions[viewinds[i]])
ax = f.add_subplot(num_views,2,i*2+1)
ax.bar(xvals,np.round(tempY*60))
#ax.set_title('training data originals')
ax = f.add_subplot(num_views,2,i*2+2)
ax.bar(xvals,np.round(tempY2*60))
plt.show()
#c=vae_all.transform(train_originals)
#print("New representation", c.shape, c[0])
_,_,quants=vae_all(train_originals,is_training=False)
quant_t,quant_b = quants
qt,qb = quant_t["encoding_indices"].numpy(),quant_b["encoding_indices"].numpy()
print("Int Representations")
for t,b in zip(qt,qb):
print(t,b)
#print("Int Representations",quant_b["encoding_indices"].shape,quant_b["encoding_indices"])
#print("Int Representations",quant_t["encoding_indices"].shape,quant_t["encoding_indices"])