def test_soft_clip(self):
wav = np.array([-1, -0.5, 0, 0.5, 1])
assert np.allclose(
transforms.SoftClip()(wav),
np.array([0.26894142, 0.37754067, 0.5, 0.62245933, 0.73105858]))
assert np.allclose(
transforms.SoftClip(1)(wav),
np.array([0.26894142, 0.37754067, 0.5, 0.62245933, 0.73105858]))
assert np.allclose(
transforms.SoftClip(1.)(wav),
np.array([0.26894142, 0.37754067, 0.5, 0.62245933, 0.73105858]))
with pytest.raises(ValueError):
transforms.SoftClip('a')
from torch.nn import CrossEntropyLoss
from torch.utils.data import DataLoader
from torch.utils.data import random_split
import yews.datasets as dsets
import yews.transforms as transforms
from yews.models import Cpic
from yews.train import Trainer
if __name__ == '__main__':
# Preprocessing
waveform_transform = transforms.Compose([
transforms.ZeroMean(),
transforms.SoftClip(1e-4),
transforms.ToTensor(),
])
# Prepare dataset
dset = dsets.Wenchuan(path='.', download=True,
sample_transform=waveform_transform)
# Split datasets into training and validation
train_length = int(len(dset) * 0.8)
val_length = len(dset) - train_length
train_set, val_set = random_split(dset, [train_length, val_length])
# Prepare dataloaders
train_loader = DataLoader(train_set, batch_size=100, shuffle=True, num_workers=4)
val_loader = DataLoader(val_set, batch_size=1000, shuffle=False, num_workers=8)
# from yews.models import polarity_v1
# polarity=polarity_v1
from yews.models import fm_v1
from yews.models import fm_v2
focal_mechanism = fm_v2
if __name__ == '__main__':
print("Now: start : " + str(datetime.datetime.now()))
# Preprocessing
waveform_transform = transforms.Compose([
transforms.ZeroMean(),
transforms.SoftClip(1e-2),
transforms.ToTensor(),
])
# Prepare dataset
dsets.set_memory_limit(10 * 1024**3) # first number is GB
# dset = dsets.Wenchuan(path='/home/qszhai/temp_project/deep_learning_course_project/cpic', download=False,sample_transform=waveform_transform)
# dset = dsets.SCSN_polarity(path='/home/qszhai/temp_project/deep_learning_course_project/first_motion_polarity/scsn_data/npys', download=False, sample_transform=waveform_transform)
# dset = dsets.Taiwan_focal_mechanism(path='/home/qszhai/temp_project/deep_learning_course_project/focal_mechanism/npys_for_focal_mechanism', download=False, sample_transform=waveform_transform)
# # Split datasets into training and validation
# train_length = int(len(dset) * 0.8)
# val_length = len(dset) - train_length
# train_set, val_set = random_split(dset, [train_length, val_length])
train_set = dsets.Taiwan_focal_mechanism(
def test_soft_clip(self):
wav = np.array([-1, -0.5, 0, 0.5, 1])
assert np.allclose(transforms.SoftClip()(wav),
np.array([0.26894142, 0.37754067, 0.5, 0.62245933, 0.73105858]))
from yews.train import Trainer
#from yews.models import cpic
from yews.models import cpic_v1
from yews.models import cpic_v2
from yews.models import cpic_v3
cpic = cpic_v3
if __name__ == '__main__':
print("Now: start : " + str(datetime.datetime.now()))
# Preprocessing
waveform_transform = transforms.Compose([
transforms.ZeroMean(),
transforms.SoftClip(1e-3),
transforms.ToTensor(),
])
# Prepare dataset
dsets.set_memory_limit(10 * 1024**3) # first number is GB
dset = dsets.Wenchuan(
path=
'/home/qszhai/temp_project/deep_learning_course_project/cpic/wenchuan_data/train_npy',
download=False,
sample_transform=waveform_transform)
# Split datasets into training and validation
train_length = int(len(dset) * 0.8)
val_length = len(dset) - train_length
train_set, val_set = random_split(dset, [train_length, val_length])