-
Notifications
You must be signed in to change notification settings - Fork 3
/
make_test_dataset.py
37 lines (30 loc) · 1.56 KB
/
make_test_dataset.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
import recovar.config
from importlib import reload
from recovar import simulator, output, utils
reload(simulator)
import numpy as np
import os
# atom_coeff_path = 'data/atom_coeffs_extended.json'
# with open(os.path.join(os.path.dirname(__file__), atom_coeff_path), 'r') as f:
# atom_coeffs = json.load(f)
def make_test_dataset(noise_level = 0.1, n_images = None):
grid_size =64
this_dir = os.path.dirname(__file__)
volume_folder_input = this_dir+ '/recovar/data/vol'
print(volume_folder_input)
output_folder = this_dir + '/test_dataset/'
output.mkdir_safe(output_folder)
outlier_file_input = None
log_n = 3
n_images = int(10**(log_n)) if n_images is None else n_images
voxel_size = 4.25 * 128 / grid_size
volume_distribution = np.array([1/4, 1/4, 1/2])
image_stack, sim_info = simulator.generate_synthetic_dataset(output_folder, voxel_size, volume_folder_input, n_images,
outlier_file_input = outlier_file_input, grid_size = grid_size,
volume_distribution = volume_distribution, dataset_params_option = "uniform", noise_level =noise_level,
noise_model = "radial1", put_extra_particles = False, percent_outliers = 0.00,
volume_radius = 0.7, trailing_zero_format_in_vol_name = True, noise_scale_std = 0.2 * 0, contrast_std =0.1 , disc_type = 'linear_interp')
print(f"Finished generating dataset {output_folder}")
if __name__ == '__main__':
make_test_dataset()
print("Done")