plt.xlabel('Time Symbol')
plt.ylabel(' Original Req')
plt.savefig('results' '/Main' + '.png', dpi=700)
plt.pause(3)
plt.close()
if __name__ == '__main__':
global_start_time = time.time()
epochs = 500
seq_len = 25
X_train, y_train,y_train_original_part, X_test, y_test,ts_train,ts_test,minMaxScaler = \
Train_LSTM.load_data(seq_len)
model = Train_LSTM.build_model([1, seq_len, 50, 10, 1])
#from keras.utils.vis_utils import plot_model
#plot_model(model, to_file='model_plot.png', show_shapes=True, show_layer_names=True)
print(np.array(X_train).shape)
print(np.array(y_train).shape)
print('> Data Loaded. Compiling...')
st1 = time.time()
history = model.fit(X_train,
y_train,
batch_size=128,
nb_epoch=epochs,
validation_split=0.1)
+ '.png',
dpi=700)
plt.pause(3)
plt.close()
if __name__ == '__main__':
global_start_time = time.time()
epochs = 100
seq_len = 25
factor = 0.8
mode = 2 ## 1 for CPU, 2 for RAM
X_train, y_train,y_train_original_part, X_test, y_test,ts_train,ts_test,min_max_scaler = \
Train_LSTM.load_data(seq_len,mode,factor,first_plot=True)
model = Train_LSTM.build_model([1, seq_len, 50, 1])
#from keras.utils.vis_utils import plot_model
#plot_model(model, to_file='model_plot.png', show_shapes=True, show_layer_names=True)
print(np.array(X_train).shape)
print(np.array(y_train).shape)
print('> Data Loaded. Compiling...')
st1 = time.time()
history = model.fit(X_train,
y_train,
batch_size=512,
nb_epoch=epochs,
validation_split=0.1)
plt.legend()
plt.grid()
plt.savefig('results' '/Main' + '.png', dpi=700)
plt.pause(3)
plt.close()
if __name__ == '__main__':
global_start_time = time.time()
#epochs = 50
seq_len = 25
X_train, y_train,y_train_original_part, X_test, y_test,ts_train,ts_test,minMaxScaler = \
Train_LSTM.load_data(seq_len)
# model = Train_LSTM.build_model([1,seq_len, 50,10, 1])
# #from keras.utils.vis_utils import plot_model
# #plot_model(model, to_file='model_plot.png', show_shapes=True, show_layer_names=True)
#
# print(np.array(X_train).shape)
# print(np.array(y_train).shape)
#
# print('> Data Loaded. Compiling...')
# st1=time.time()
# history=model.fit(
# X_train,
# y_train,
# batch_size=128,
# nb_epoch=epochs,
plt.pause(5)
plt.close()
print('writing to DB!')
print(len(ts_test), len(ypr_revert), len(ytr_revert), len(X_test))
for k in range(len(ts_test)):
#print(ts_test[k], ytr_revert[k], ypr_revert[k],ytr[k])
cur.execute('update nasa_http_emd_1min_copy set num_req_pred_gan=%s where imf_index=%s'
' and num_req_pred is null and ts=%s', \
(float(ypr_revert[k]), int(imf_index), int(ts_test[k])+seq_len+1))
conn.commit()
if __name__ == '__main__':
seq_len = 30
norm_version = 1 # v2= MinMaxScaler(0,1) , v1=MaxAbsScaler(-1,1)
for imf_index in range(1, 4):
X_train, y_train, y_train_original_part, X_test, y_test, ts_train, ts_test, MaxAbsScalerObj = \
Train_LSTM.load_data(seq_len, imf_index, norm_version)
print(' --------------\n Shape of data is : \n ')
print('X_train: ', X_train.shape, ' Y_train: ', y_train.shape)
print('X_test: ', X_test.shape, ' Y_test: ', y_test.shape)
print('----------------\n')
gan = GAN()
gan.train(epochs=50, batchsize=64, verbose=False)
gan.test()
cur.execute('insert into calgary_http_emd_60min_copy (ts,num_of_req,imf_index,num_req_pred) values(%s,%s,%s,%s) ',
(int(ts_test[k]),int(y_test[k]),imf,float(y_pred[k])))
conn.commit()
if __name__=='__main__':
for ii in range(1,18):
global_start_time = time.time()
imf_index = ii
epochs = 500 if ii<5 else 200
seq_len = 10
norm_version=1 # v2= MinMaxScaler(0,1) , v1=MaxAbsScaler(-1,1)
X_train, y_train,y_train_original_part, X_test, y_test,ts_train,ts_test,MaxAbsScalerObj =\
Train_LSTM.load_data(seq_len,imf_index,norm_version)
model = Train_LSTM.build_model([1, seq_len, 20,1])
from keras.utils.vis_utils import plot_model
#plot_model(model, to_file='model_plot.png', show_shapes=True, show_layer_names=True)
print(np.array(X_train).shape)
print(np.array(y_train).shape)
print('> Data Loaded. Compiling...')
history =model.fit(
X_train,
y_train,
label='Prediction Real Data, MAPE = %.4f%% ,\n '
' RMSE=%.4f , RMSRE=%.4f ' %
(map_denormalize, rms_denormalize, rmsre_denorm))
plt.legend()
plt.savefig(
'/home/vacek/Cloud/cloud-predictor/NASA-HTTP/prediction/GANS-only/10min/resutls'
+ '/prediction_original' + '.png',
dpi=700)
plt.pause(5)
plt.close()
if __name__ == '__main__':
seq_len = 30
norm_version = 2 # v2= MinMaxScaler(0,1) , v1=MaxAbsScaler(-1,1)
X_train, y_train, y_train_original_part, X_test, y_test, ts_train, ts_test, MaxAbsScalerObj = \
Train_LSTM.load_data(seq_len, norm_version)
print(' --------------\n Shape of data is : \n ')
print('X_train: ', X_train.shape, ' Y_train: ', y_train.shape)
print('X_test: ', X_test.shape, ' Y_test: ', y_test.shape)
print('----------------\n')
gan = GAN()
gan.train(epochs=50, batchsize=32, verbose=False)
gan.test()
ax = fig.add_subplot(111)
ax.plot(true_data, label='True Data')
#Pad the list of predictions to shift it in the graph to it's correct start
for i, data in enumerate(predicted_data):
padding = [None for p in range(i * prediction_len)]
plt.plot(padding + data, label='Prediction-cs-200-first-10-part')
plt.legend()
plt.show()
if __name__=='__main__':
global_start_time = time.time()
epochs = 40
seq_len = 50
mode=2 ## 1 for CPU, 2 for RAM
X_train, y_train,y_train_original_part, X_test, y_test,ts_train,ts_test = Train_LSTM.load_data(seq_len,mode)
model = Train_LSTM.build_model([1, 50, 100, 1])
print(np.array(X_train).shape)
print(np.array(y_train).shape)
print('> Data Loaded. Compiling...')
model.fit(
X_train,
y_train,
batch_size=512,
nb_epoch=epochs,
validation_split=0.05)
# plt.xlabel('Time Symbol')
# plt.ylabel('Normalized CPU Req')
#
plt.savefig('RAM.png', format='png', dpi=800)
plt.show()
if __name__ == '__main__':
global_start_time = time.time()
epochs = 50
seq_len = 25
factor = 0.8
mode = 2 ## 1 for CPU, 2 for RAM
X_train, y_train,y_train_original_part, X_test, y_test,ts_train,ts_test = \
Train_LSTM.load_data(seq_len,mode,factor,first_plot=True)
from keras.models import load_model
if mode == 1:
model = load_model('model-CPU.h5')
elif mode == 2:
model = load_model('model-RAM.h5')
predicted = Train_LSTM.predict_point_by_point(model, X_test)
print(len(predicted), len(y_test), '------------')
del X_train, X_test, y_train
print('-----\n--------------\n--------------------------')
sleep(3)
print('Training duration (s) : ', time.time() - global_start_time)
ax = fig.add_subplot(313)
plt.plot(ts_test, predicted_data, color='green', label='Prediction')
plt.legend()
plt.grid()
plt.show()
if __name__=='__main__':
global_start_time = time.time()
epochs = 10
seq_len =32
factor=0.8
mode=1 ## 1 for CPU, 2 for RAM
X_train, y_train,y_train_original_part, X_test, y_test,ts_train,ts_test = Train_LSTM.load_data(seq_len,mode,factor,first_plot=True)
model = Train_LSTM.build_model([1,32, 64, 1])
print(np.array(X_train).shape)
print(np.array(y_train).shape)
print('> Data Loaded. Compiling...')
history=model.fit(
X_train,
y_train,
batch_size=64,
nb_epoch=epochs,
validation_split=0.2)