-
Notifications
You must be signed in to change notification settings - Fork 0
/
test_bootstrap.py
25 lines (16 loc) · 951 Bytes
/
test_bootstrap.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
import h5py
from keras.utils.io_utils import HDF5Matrix
import IPython
import numpy as np
def normalize_pixel(data):
return data/255-.5
y4_data = HDF5Matrix('/home/ubuntuone/Projects/data/CelebAHDF5/celeba_aligned_cropped_train_5cls_bootstrap4.hdf5', 'targets')
y2_data = HDF5Matrix('/home/ubuntuone/Projects/data/CelebAHDF5/celeba_aligned_cropped_train_5cls_bootstrap2.hdf5', 'targets')
y1_data = HDF5Matrix('/home/ubuntuone/Projects/data/CelebAHDF5/celeba_aligned_cropped_train_5cls_bootstrap1.hdf5', 'targets')
y3_data = HDF5Matrix('/home/ubuntuone/Projects/data/CelebAHDF5/celeba_aligned_cropped_train_5cls_bootstrap3.hdf5', 'targets')
y0_data = HDF5Matrix('/home/ubuntuone/Projects/data/CelebAHDF5/celeba_aligned_cropped_train_5cls_bootstrap0.hdf5', 'targets')
y_data = y4_data[:]
cnt = np.where(y_data[:,2]==1)[0].shape[0]
# sample = np.random.random_integer(0,)
# y_data = np.concatenate((y_data, y2_data[sample]))
IPython.embed()