# Build the data generators
if load_memory or use_keras_fit:
print('Loading data to memory...')
generator = SeriesDataGenerator(dlwp,
train_data,
input_sel=input_selection,
output_sel=output_selection,
input_time_steps=input_time_steps,
output_time_steps=output_time_steps,
batch_size=batch_size,
add_insolation=add_solar,
load=load_memory,
shuffle=shuffle,
interval=step_interval)
if use_keras_fit:
p_train, t_train = generator.generate([])
if validation_data is not None:
val_generator = SeriesDataGenerator(dlwp,
validation_data,
input_sel=input_selection,
output_sel=output_selection,
input_time_steps=input_time_steps,
output_time_steps=output_time_steps,
batch_size=batch_size,
add_insolation=add_solar,
load=load_memory,
interval=step_interval)
if use_keras_fit:
val = val_generator.generate([])
else:
val_generator = None
dlwp.time_dim = 2
sequence = dlwp._n_steps
else:
sequence = None
# Create data generator
generator = SeriesDataGenerator(dlwp,
data,
batch_size=216,
input_sel=input_selection,
output_sel=output_selection,
input_time_steps=input_time_steps,
output_time_steps=output_time_steps,
add_insolation=add_insolation,
sequence=sequence)
p_val, t_val = generator.generate([], scale_and_impute=False)
# Create TimeSeriesEstimator
estimator = TimeSeriesEstimator(dlwp, generator)
# Make a time series prediction and convert the predictors for comparison
print('Predicting with model %s...' % model_label)
forecast = estimator.predict(num_plot_steps, verbose=1)
forecast = forecast.sel(**selection)
# Scale the forecast
if scale_variables:
forecast = forecast * variable_std + variable_mean
# Take the Laplacian if vorticity is desired
if plot_laplace:
# Build the data generators
if load_memory or use_keras_fit:
print('Loading data to memory...')
generator = SeriesDataGenerator(dlwp,
train_data,
input_sel=io_selection,
output_sel=io_selection,
input_time_steps=io_time_steps,
output_time_steps=io_time_steps,
sequence=integration_steps,
add_insolation=add_solar,
batch_size=batch_size,
load=load_memory,
shuffle=shuffle)
if use_keras_fit:
p_train, t_train = generator.generate([])
if validation_data is not None:
val_generator = SeriesDataGenerator(dlwp,
validation_data,
input_sel=io_selection,
output_sel=io_selection,
input_time_steps=io_time_steps,
output_time_steps=io_time_steps,
sequence=integration_steps,
add_insolation=add_solar,
batch_size=batch_size,
load=load_memory)
if use_keras_fit:
val = val_generator.generate([])
else:
val_generator = None
weighting='midlatitude')
else:
loss_function = 'mse'
# Build the model
dlwp.build_model(layers, loss=loss_function, optimizer='adam', metrics=['mae'])
print(dlwp.model.summary())
#%% Initialize the scaler/imputer if necessary
# Generate the data to fit the scaler and imputer. The generator will by default apply scaling because it is necessary
# to automate its use in the Keras fit_generator method, so disable it when dealing with data to fit the scaler and
# imputer. If using pre-scaled data (i.e., dlwp was initialized with scaler_type=None and the Preprocessor data was
# generated with scale_variables=True), this step should be omitted.
fit_set = list(range(2)) # Use a much larger value when it matters
p_fit, t_fit = generator.generate(fit_set, scale_and_impute=False)
dlwp.init_fit(p_fit, t_fit)
p_fit, t_fit = (None, None)
#%% Train, evaluate, and save the model
# Train and evaluate the model
start_time = time.time()
print('Begin training...')
history = History()
early = EarlyStoppingMin(
min_epochs=min_epochs,
monitor='val_loss' if val_generator is not None else 'loss',
min_delta=0.,
patience=patience,
restore_best_weights=True,
# Build the data generators
if load_memory or use_keras_fit:
print('Loading data to memory...')
generator = SeriesDataGenerator(dlwp,
train_data,
input_sel=io_selection,
output_sel=io_selection,
input_time_steps=io_time_steps,
output_time_steps=io_time_steps,
sequence=integration_steps,
add_insolation=add_solar,
batch_size=batch_size,
load=load_memory,
shuffle=shuffle)
if use_keras_fit:
p_train, t_train = generator.generate([])
if validation_data is not None:
val_generator = SeriesDataGenerator(dlwp,
validation_data,
input_sel=io_selection,
output_sel=io_selection,
input_time_steps=io_time_steps,
output_time_steps=io_time_steps,
sequence=integration_steps,
add_insolation=add_solar,
batch_size=batch_size,
load=load_memory)
if use_keras_fit:
val = val_generator.generate([])
else:
val_generator = None