callbacks,
epoch_size=20,
epoch_steps=100)
ml.save_model(model,'entersolar_model.h5')
output = ml.load_models('entersolar_model.h5')
import importlib
importlib.reload(ml)
start_idx = 0
length = 2000
target_names = target_output
y_pred, y_true,x,mse = ml.plot_comparison(x_test_scaled,y_test,y_scaler,
target_names,output,
start_idx = start_idx,
length=length,
verbose=True)
solar_pvwatts = helper.pv_data(coordinates=(lat,lon),solar_size_kw_dc = 1750,tilt=25)
index = pd.date_range(start='1/1/'+str(2018), periods=8760, freq='60T')
solar_pvwatts.index = index
solar_pvwatts = solar_pvwatts.resample('15T').asfreq()
#solar_pvwatts = solar_pvwatts.drop(columns={'month','day','hour'})
solar_pvwatts = solar_pvwatts.interpolate()
target_output,
batch_size=256,
sequence_length=24,
training_len=0.75,
validation_len=0.225)
model = ml.train_model(model,
batch,
validation_data,
x_test_scaled,
y_test_scaled,
callbacks,
epoch_size=25,
epoch_steps=100)
ml.save_model(model, 'model.h5')
output = ml.load_models('model.h5')
import importlib
importlib.reload(ml)
start_idx = 0
length = 500
target_names = target_output
y_pred, y_true, x = ml.plot_comparison(x_test_scaled,
y_test,
y_scaler,
target_names,
output,
length=length)
callback_reduce_lr]
model.fit_generator(generator=batch,
epochs=15,
steps_per_epoch=100,
validation_data=validation_data,
callbacks=callbacks)
result = model.evaluate(x=np.expand_dims(x_test_scaled, axis=0),
y=np.expand_dims(y_test_scaled, axis=0))
start_idx = 0
length = 10000
target_names = target_output
y_pred, y_true,x = ml.plot_comparison(x_test_scaled,y_test,y_scaler,
target_names,model)
mse_vect = list()
for i in range(len(y_pred)):
mse_vect.append((y_pred[i]-y_true[i])**2)
mse = sum(mse_vect)
plt.figure()
plt.plot(mse_vect)
data_size = 144
training_size = 144
new = np.array(data.solar_irradiance).reshape(8760,2)
new_expand = np.expand_dims(new,axis=0)