Skip to content

Commit

Permalink
Update talos
Browse files Browse the repository at this point in the history
  • Loading branch information
@jacobbieker
jacobbieker committed Feb 10, 2019
1 parent 13985d4 commit 2c591cf
Show file tree
Hide file tree
Showing 4 changed files with 182 additions and 76 deletions.
98 changes: 57 additions & 41 deletions examples/talos_flat_classification.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -7,33 +7,56 @@
from talos.model.early_stopper import early_stopper
from talos.model.layers import hidden_layers
from talos.model.normalizers import lr_normalizer
from talos.metrics.keras_metrics import root_mean_squared_error, fmeasure_acc, matthews_correlation_acc, precision_acc, recall_acc
from talos.metrics.keras_metrics import root_mean_squared_error, fmeasure_acc, matthews_correlation_acc, precision_acc, \
recall_acc

from factnn.utils.cross_validate import get_chunk_of_data
from factnn.utils.cross_validate import get_chunk_of_data, get_data_generators

# Get arguments


import argparse

# construct the argument parse and parse the arguments
ap = argparse.ArgumentParser()
ap.add_argument("-d", "--dir", required=True,
help="Directory for Files")
ap.add_argument("-s", "--size", type=int, required=True,
help="Chunk Size")
ap.add_argument("-g", "--grid", type=float, required=True,
help="Downsample Amount")

args = vars(ap.parse_args())

directory = args['dir'] # "/media/jacob/WDRed8Tb1/Insync/iact_events/"
gamma_dir = [directory + "gammaFeature/clump20/"]
proton_dir = [directory + "protonFeature/clump20/"]

# Parameter Dictionary for talos

params = {'lr': (1, 10, 10),
'first_neuron': [16, 32, 64],
'last_neuron': [8, 16, 32],
params = {'lr': (1, 10, 5),
'first_neuron': [16, 64],
'last_neuron': [8, 32],
'hidden_layers': [2, 3, 6],
'batch_size': [2, 16, 64],
'batch_size': [2, 64],
'epochs': [500],
'dropout': (0, 0.80, 4),
'dropout': (0, 0.80, 3),
'weight_regulizer': [None],
'emb_output_dims': [None],
'optimizer': [adam, nadam, rmsprop],
'losses': [mean_squared_error],
'activation': [relu, elu, hard_sigmoid],
'last_activation': [softmax],

'neuron_1': [8, 16, 64],
'kernel_1': [1, 3, 5],
'neuron_1': [8, 64],
'kernel_1': [1, 3],
'stride_1': [1, 2, 3],
'layer_drop': [0.0, 0.8, 4],
'layers': [2,3,4],
'second_conv': [0,1],
'pool': [0,1]
'layer_drop': [0.0, 0.8, 3],
'layers': [2, 3, 4],
'second_conv': [0, 1],
'pool': [0, 1],
'rebin': [25, 50, 75, 100],
'time': [1, 2, 3]

}
'''
Expand All @@ -49,12 +72,17 @@
'stride_3': [1, 2],
'''


def input_model(x_train, y_train, x_val, y_val, params):
train_gen, val_gen, _, _ = get_data_generators(directory=gamma_dir, proton_directory=proton_dir,
indicies=(30, 129, params['time']), rebin=params['rebin'],
batch_size=params['batch_size'], as_channels=True,
max_elements=args['size'])
model = Sequential()

model.add(Conv2D(params['neuron_1'], kernel_size=params['kernel_1'], strides=params['stride_1'],
padding='same',
input_shape=(100, 100, 1),
input_shape=(params['rebin'], params['rebin'], params['time']),
activation=params['activation']))
if params['second_conv']:
model.add(Conv2D(params['neuron_1'], kernel_size=params['kernel_1'], strides=1,
Expand All @@ -68,7 +96,7 @@ def input_model(x_train, y_train, x_val, y_val, params):
if params['second_conv']:
model.add(Conv2D(params['neuron_1'], kernel_size=params['kernel_1'], strides=1,
padding='same', activation=params['activation']))
#if params['pool']:
# if params['pool']:
# model.add(MaxPooling2D())
if params['layer_drop'] > 0.001:
model.add(Dropout(params['layer_drop']))
Expand Down Expand Up @@ -108,42 +136,30 @@ def input_model(x_train, y_train, x_val, y_val, params):
precision_acc,
recall_acc])

out = model.fit(x_train, y_train,
batch_size=params['batch_size'],
epochs=params['epochs'], verbose=0,
validation_data=[x_val, y_val],
callbacks=[early_stopper(params['epochs'],
mode='moderate')])
out = model.fit_generator(
generator=train_gen,
epochs=params['epochs'],
verbose=0,
validation_data=val_gen,
callbacks=[early_stopper(params['epochs'],
mode='moderate')],
use_multiprocessing=True,
workers=10,
max_queue_size=50,
)

return out, model


import argparse

# construct the argument parse and parse the arguments
ap = argparse.ArgumentParser()
ap.add_argument("-d", "--dir", required=True,
help="Directory for Files")
ap.add_argument("-s", "--size", type=int, required=True,
help="Chunk Size")
ap.add_argument("-g", "--grid", type=float, required=True,
help="Downsample Amount")

args = vars(ap.parse_args())

directory = args['dir'] # "/media/jacob/WDRed8Tb1/Insync/iact_events/"
gamma_dir = [directory + "gammaFeature/no_clean/"]
proton_dir = [directory + "protonFeature/no_clean/"]

x, y = get_chunk_of_data(directory=gamma_dir, proton_directory=proton_dir, indicies=(30, 129, 1), rebin=100,
chunk_size=args['size'], as_channels=True)
x, y = get_chunk_of_data(directory=gamma_dir, proton_directory=proton_dir, indicies=(30, 129, 1), rebin=25,
chunk_size=2, as_channels=True)
print("Got data")
print("X Shape", x.shape)
print("Y Shape", y.shape)
history = ta.Scan(x, y,
params=params,
dataset_name='flat_separation_test',
experiment_no='1',
experiment_no='2',
model=input_model,
search_method='random',
grid_downsample=args['grid'])
81 changes: 46 additions & 35 deletions examples/talos_time_classification.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -10,32 +10,51 @@
from talos.metrics.keras_metrics import root_mean_squared_error, fmeasure_acc, matthews_correlation_acc, precision_acc, recall_acc


from factnn.utils.cross_validate import get_chunk_of_data
from factnn.utils.cross_validate import get_chunk_of_data, get_data_generators


import argparse

# construct the argument parse and parse the arguments
ap = argparse.ArgumentParser()
ap.add_argument("-d", "--dir", required=True,
help="Directory for Files")
ap.add_argument("-s", "--size", type=int, required=True,
help="Chunk Size")
ap.add_argument("-g", "--grid", type=float, required=True,
help="Downsample Amount")

args = vars(ap.parse_args())

directory = args['dir'] # "/media/jacob/WDRed8Tb1/Insync/iact_events/"
gamma_dir = [directory + "gammaFeature/clump20/"]
proton_dir = [directory + "protonFeature/clump20/"]

# Parameter Dictionary for talos

params = {'lr': (1, 10, 10),
'first_neuron': [4, 16, 32],
params = {'lr': (1, 10, 5),
'first_neuron': [4, 32],
'last_neuron': [4, 16],
'hidden_layers': [2, 4],
'batch_size': [2, 8],
'epochs': [500],
'dropout': (0, 0.80, 4),
'dropout': (0, 0.80, 3),
'weight_regulizer': [None],
'emb_output_dims': [None],
'optimizer': [adam, nadam, rmsprop],
'losses': [categorical_crossentropy, logcosh],
'activation': [relu, elu],

'neuron_1': [8, 16, 64],
'neuron_1': [8, 16, 32],
'kernel_1': [1, 3, 5],
'stride_1': [1, 2, 3],
'rec_dropout': [0.0, 0.8, 4],
'rec_dropout': [0.0, 0.8, 3],
'rec_act': [hard_sigmoid, tanh],
'layer_drop': [0.0, 0.8, 4],
'layer_drop': [0.0, 0.8, 3],
'layers': [2,3,4],
'pool': [0, 1],

'rebin': [25, 50, 75, 100],
'time': [10, 15, 20]
}
'''
Expand All @@ -49,6 +68,11 @@
'''

def input_model(x_train, y_train, x_val, y_val, params):
train_gen, val_gen, _, _ = get_data_generators(directory=gamma_dir, proton_directory=proton_dir,
indicies=(30, 129, params['time']), rebin=params['rebin'],
batch_size=params['batch_size'], as_channels=False,
max_elements=args['size'])

model = Sequential()

model.add(ConvLSTM2D(params['neuron_1'], kernel_size=params['kernel_1'], strides=params['stride_1'],
Expand Down Expand Up @@ -120,43 +144,30 @@ def input_model(x_train, y_train, x_val, y_val, params):
precision_acc,
recall_acc])

out = model.fit(x_train, y_train,
batch_size=params['batch_size'],
epochs=params['epochs'], verbose=0,
validation_data=[x_val, y_val],
callbacks=[early_stopper(params['epochs'],
mode='moderate')])
out = model.fit_generator(
generator=train_gen,
epochs=params['epochs'],
verbose=0,
validation_data=val_gen,
callbacks=[early_stopper(params['epochs'],
mode='moderate')],
use_multiprocessing=True,
workers=10,
max_queue_size=50,
)

return out, model


import argparse

# construct the argument parse and parse the arguments
ap = argparse.ArgumentParser()
ap.add_argument("-d", "--dir", required=True,
help="Directory for Files")
ap.add_argument("-s", "--size", type=int, required=True,
help="Chunk Size")
ap.add_argument("-g", "--grid", type=float, required=True,
help="Downsample Amount")

args = vars(ap.parse_args())

directory = args['dir'] # "/media/jacob/WDRed8Tb1/Insync/iact_events/"
gamma_dir = [directory + "gammaFeature/no_clean/"]
proton_dir = [directory + "protonFeature/no_clean/"]

x, y = get_chunk_of_data(directory=gamma_dir, proton_directory=proton_dir, indicies=(30, 129, 10), rebin=100,
chunk_size=args['size'], as_channels=False)
x, y = get_chunk_of_data(directory=gamma_dir, proton_directory=proton_dir, indicies=(30, 129, 10), rebin=25,
chunk_size=2, as_channels=False)

print("Got data")
print("X Shape", x.shape)
print("Y Shape", y.shape)
history = ta.Scan(x, y,
params=params,
dataset_name='time_separation_test',
experiment_no='1',
experiment_no='2',
model=input_model,
search_method='random',
grid_downsample=args['grid'])
1 change: 1 addition & 0 deletions examples/test_event_files.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -83,6 +83,7 @@ def d(clump_size, path):


if __name__ == '__main__':

clump_size = 20
output_paths = [os.path.join(output_path, "protonFeature", "no_clean"),os.path.join(output_path, "protonFeature", "clump"+str(clump_size)),os.path.join(output_path, "protonFeature", "core"+str(clump_size)),
os.path.join(output_path, "gammaFeature", "no_clean"),os.path.join(output_path, "gammaFeature", "clump"+str(clump_size)),os.path.join(output_path, "gammaFeature", "core"+str(clump_size)),
Expand Down

0 comments on commit 2c591cf

Please sign in to comment.