width = np.random.uniform(0.1, 2)
pnum = 10
amplitude = np.random.uniform(level * 2 / 3, level)
signal = stg.generate(ts,
fs,
stg.constant_path(f_start=fs[start_index],
drift_rate=drift_rate),
stg.periodic_gaussian_t_profile(
period, phase, sigma, pulse_dir, width, pnum,
amplitude, level),
stg.gaussian_f_profile(width=line_width),
stg.constant_bp_profile(level=1.0),
integrate=True)
signal = stg.normalize(stg.inject_noise(signal),
cols=128,
exclude=0.2,
use_median=False)
plt.imsave(output_fn, signal)
print('Saved %s of %s scintillated data for %s' % (i + 1, num, name))
for i in range(num):
output_fn = '/datax/scratch/bbrzycki/data/%s/%s/%s/%s_%04d.png' % (
dir, name, 'constant', 'constant', i)
start_index = np.random.randint(0, fchans)
drift_rate = np.random.uniform(-start_index * df / (tsamp * tchans),
(fchans - 1 - start_index) * df /
except OSError as e:
if e.errno != errno.EEXIST:
raise
# Make csv to save data
csv_fn = '%s/%s' % (prefix, 'labels.csv')
# Generate training and validation data!
datasets = [('train', 5000), ('validation', 500)]
with open(csv_fn, 'w') as f:
writer = csv.writer(f)
for label in labels:
for process, num in datasets:
for i in range(num):
output_prefix = '%s/%s/%s/%s_%04d' % (prefix, process, label, label, i,)
##############################################################################
# Generate signals -- these parameters should be tuned to whatever makes sense
signal, [start_index, drift_rate, line_width, level, spread] = generate_training0_signal(label=label)
# Normalize and write data
normalized_signal = stg.normalize(stg.inject_noise(signal), cols = 128, exclude = 0.2, use_median=False)
plt.imsave(output_prefix + '.png', normalized_signal)
np.save(output_prefix + '.npy', normalized_signal)
writer.writerow([output_prefix, start_index, drift_rate, line_width, level, spread])
print([output_prefix, start_index, drift_rate, line_width, level, spread])
print('Saved %s of %s signal data for %s, %s' % (i + 1, num, label, process))