from blackfox import BlackFox
from blackfox import PredictionFileConfig
from blackfox import Range
blackfox_url = 'http://localhost:50476/'
bf = BlackFox(blackfox_url)
config = PredictionFileConfig(
input_ranges=[
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1)
],
output_ranges=[Range(min=0, max=1),
Range(min=0, max=1)])
bf.predict_from_file_keras(config, 'data/optimized_network_cancer.h5',
'data/cancer_test_set_input.csv',
'data/cancer_predict.csv')
from blackfox import BlackFox
from blackfox import KerasOptimizationConfig
from blackfox import OptimizationEngineConfig
from blackfox import Range
blackfox_url = 'http://localhost:50476/'
bf = BlackFox(blackfox_url)
# custom config
engine_config = OptimizationEngineConfig(
crossover_distribution_index=20,
crossover_probability=0.9,
mutation_distribution_index=20,
mutation_probability=0.01,
proc_timeout_miliseconds=200000,
max_num_of_generations=10,
population_size=20
)
config = KerasOptimizationConfig(
dropout=Range(min=0, max=25),
batch_size=10,
input_ranges=[
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
import seaborn as sns
import matplotlib as mpl
mpl.style.use('ggplot')
import warnings
warnings.filterwarnings("ignore")
from sklearn.model_selection import train_test_split
from sklearn.metrics import confusion_matrix
from sklearn.preprocessing import StandardScaler, MinMaxScaler
import h5py
import time
# blackfox_url = 'http://147.91.204.14:32702' # BF_URL
# blackfox_url = 'http://147.91.204.85:32723'# ~VODAFONE_URL
blackfox_url = 'http://147.91.204.14:32706' # BF FOR MAPE verziju blackfox 3.2.0
# blackfox_url = 'http://localhost:5000'
bf = BlackFox(blackfox_url)
dataframe = pd.read_csv(
'F:\\.spyder-py3\\projekti\\Vodena_Primeri\\ImplementingEncodingInBf\\data\\data_set.csv'
)
y = dataframe.iloc[:, -1:].values
X = dataframe.iloc[:, :-1].values
X_train, X_test, y_train, y_test = train_test_split(X,
y,
test_size=0.2,
random_state=0)
# X_train, X_val, y_train, y_val = train_test_split(X_train, y_train, test_size = 0.2, random_state = 0)
from blackfox import BlackFox
from blackfox import Range
from blackfox import KerasLayerConfig
from blackfox import KerasTrainingConfig
from blackfox import KerasHiddenLayerConfig
blackfox_url = 'http://localhost:50476/'
bf = BlackFox(blackfox_url)
input_ranges = [
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1)
]
output_layer = KerasLayerConfig(
activation_function='Sigmoid',
ranges=[Range(min=0, max=1), Range(min=0, max=1)])
hidden_layer_configs = [
KerasHiddenLayerConfig(neuron_count=6, activation_function='Sigmoid'),
KerasHiddenLayerConfig(neuron_count=6, activation_function='Sigmoid'),
KerasHiddenLayerConfig(neuron_count=6, activation_function='Sigmoid'),
KerasHiddenLayerConfig(neuron_count=6, activation_function='Sigmoid')
]
from blackfox import BlackFox
from blackfox import KerasOptimizationConfig
from blackfox import OptimizationEngineConfig
from blackfox import Range
import csv
blackfox_url = 'http://localhost:50476/'
bf = BlackFox(blackfox_url)
input_columns = 9
input_set = []
output_set = []
with open('data/cancer_training_set.csv') as csv_file:
csv_reader = csv.reader(csv_file, delimiter=',')
line_count = 0
for row in csv_reader:
if line_count == 0:
print(f'Column names are {", ".join(row)}')
else:
data = list(map(float, row))
input_set.append(data[:input_columns])
output_set.append(data[input_columns:])
line_count += 1
print(f'Processed {line_count} lines.')
# Use CTRL + C to stop optimization
network_io = bf.optimize_keras_sync(input_set, output_set)
if network_io is not None:
from blackfox import BlackFox
from blackfox import PredictionArrayConfig
from blackfox import Range
blackfox_url = 'http://localhost:50476/'
bf = BlackFox(blackfox_url)
config = PredictionArrayConfig(
data_set=[[0.2, 0.1, 0.1, 0.1, 0.2, 0.1, 0.2, 0.1, 0.1],
[0.2, 0.1, 0.1, 0.1, 0.2, 0.1, 0.3, 0.1, 0.1],
[0.5, 0.1, 0.1, 0.1, 0.2, 0.1, 0.2, 0.1, 0.1],
[0.5, 0.4, 0.6, 0.8, 0.4, 0.1, 0.8, 1, 0.1]],
input_ranges=[
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1)
],
output_ranges=[Range(min=0, max=1),
Range(min=0, max=1)])
result = bf.predict_from_array_keras(config, 'data/trained_network_cancer.h5')
print(result)
def setUp(self):
self.blackfox = BlackFox() # noqa: E501
class TestDataSetApi(unittest.TestCase):
"""Blackfox unit test stubs"""
def setUp(self):
self.blackfox = BlackFox() # noqa: E501
def tearDown(self):
pass
def test_train_keras(self):
input_layer = KerasLayerConfig(activation_function='Sigmoid',
ranges=[
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1)
])
output_layer = KerasLayerConfig(
activation_function='Sigmoid',
ranges=[Range(min=0, max=1),
Range(min=0, max=1)])
hidden_layer_configs = [
KerasHiddenLayerConfig(neuron_count=6,
activation_function='Sigmoid'),
KerasHiddenLayerConfig(neuron_count=6,
activation_function='Sigmoid'),
KerasHiddenLayerConfig(neuron_count=6,
activation_function='Sigmoid'),
KerasHiddenLayerConfig(neuron_count=6,
activation_function='Sigmoid')
]
config = KerasTrainingConfig(dropout=0,
input_layer=input_layer,
output_layer=output_layer,
hidden_layer_configs=hidden_layer_configs,
training_algorithm='Nadam',
max_epoch=3000,
cross_validation=False,
training_ratio=0.7,
random_seed=500)
self.blackfox.train_keras(
config,
'C:/Users/Korisnik/Desktop/Cancer/ulazni_podaci_cancer.csv',
'C:/Users/Korisnik/Desktop/Cancer/trained_network.onnx')
pass
def test_predict_file_keras(self):
config = PredictionFileConfig(
input_ranges=[
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1)
],
output_ranges=[Range(min=0, max=1),
Range(min=0, max=1)])
self.blackfox.predict_from_file_keras(
config, 'C:/Users/Korisnik/Desktop/Cancer/trained_network.onnx',
'C:/Users/Korisnik/Desktop/Cancer/ulazni_podaci_cancer.csv',
'C:/Users/Korisnik/Desktop/Cancer/rezultati.csv')
pass
def test_predict_array_keras(self):
config = PredictionArrayConfig(
input_ranges=[
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1)
],
output_ranges=[Range(min=0, max=1),
Range(min=0, max=1)],
data_set=[[0.50, 1.00], [0.00, 0.90], [0.85, 0.50], [0.05, 0.70],
[0.12, 1.00], [0.74, 0.05], [0.66, 0.39], [0.28, 0.11]])
results = self.blackfox.predict_from_array_keras(
config, 'C:/Users/Korisnik/Desktop/Cancer/trained_network.onnx')
print(results)
pass
def test_sha1(self):
sha1 = self.blackfox.sha1(
'C:/Users/Korisnik/Desktop/Cancer/ulazni_podaci_cancer.csv')
print(sha1)
pass
def test_upload_data_set(self):
id = self.blackfox.upload_data_set(
'C:/Users/Korisnik/Desktop/Cancer/ulazni_podaci_cancer.csv')
print(id)
pass
def test_download_data_set(self):
self.blackfox.download_data_set(
'f56e2c4fa71050ee4f55c6335947ad0b9bd47d85',
'C:/Users/Korisnik/Desktop/Cancer/ulazni_podaci_cancer_download.csv'
)
pass
def test_download_network(self):
self.blackfox.download_network(
'B9AB1F9129944A2B7D310B05B24C5CBA097A6A86',
'C:/Users/Korisnik/Desktop/Cancer/network_download.csv')
pass
def test_optimize(self):
engine_config = OptimizationEngineConfig(
crossover_distribution_index=20,
crossover_probability=0.9,
mutation_distribution_index=20,
mutation_probability=0.01,
proc_timeout_miliseconds=200000,
max_num_of_generations=50,
number_of_constraints=0,
population_size=100,
number_of_eval_per_request=1)
config = KerasOptimizationConfig(
dropout=Range(min=0, max=25),
dataset_id='f56e2c4fa71050ee4f55c6335947ad0b9bd47d85',
batch_size=10,
input_ranges=[
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1),
Range(min=0, max=1)
],
output_ranges=[Range(min=0, max=1),
Range(min=0, max=1)],
hidden_layer_count_range=Range(min=1, max=15),
neurons_per_layer=Range(min=1, max=10),
training_algorithms=[
"SGD", "RMSprop", "Adagrad", "Adadelta", "Adam", "Adamax",
"Nadam"
],
activation_functions=[
"SoftMax", "Elu", "Selu", "SoftPlus", "SoftSign", "ReLu",
"TanH", "Sigmoid", "HardSigmoid", "Linear"
],
max_epoch=3000,
cross_validation=False,
training_ratio=0.7,
random_seed=100,
engine_config=engine_config)
id = self.blackfox.optimize_keras(config)
print(id)
pass
def test_optimization_stop(self):
self.blackfox.stop_optimization_keras(
'c2db8231-da65-4821-b628-b0c8be17db7d',
'C:/Users/Korisnik/Desktop/Cancer/net.onnx')
pass
def test_optimization_status(self):
res = self.blackfox.get_optimization_status_keras(
'c2db8231-da65-4821-b628-b0c8be17db7d',
'C:/Users/Korisnik/Desktop/Cancer/net.onnx')
print(res)