Issues reproducing sine wave example

[robotdude17]

I've tried to reproduce the sine wave example using the following code in the main function seen below.

Apparently, something went wrong, see the image I have received.

Do you know what I need to fix in order to reproduce your example?

Thanks!

#Main Run Thread
if name=='main':

os.system('clear')
global_start_time = time.time()
epochs  = 1
seq_len = 50

print '> Loading data... '

X_train, y_train, X_test, y_test = lstm.load_data('sinwave.csv', seq_len, True)

print '> Data Loaded. Compiling...'

model = lstm.build_model([1, 50, 100, 1])

model.fit(
    X_train,
    y_train,
    batch_size=512,
    nb_epoch=epochs,
    validation_split=0.05)

predicted = lstm.predict_point_by_point(model, X_test)        

print 'Training duration (s) : ', time.time() - global_start_time
plot_results(predicted, y_train)

[jaungiers]

For starters it seems you've got an issue loading the actual sin wave data... The true data should be a high-frequency sin wave over the time period... Not a constant of -2.0... Please review your ETL procedures and data file.

[simonholgate]

I had a similar problem. I traced it to the importing of the sin wave data which is all quoted strings (at least on OS X High Sierra and Python 3.6.3). Converting the data after reading and decoding data = f.decode().split('\n') with:
data = list(map(float, data)) (which converts all the data to floats) to seems to do the trick.

[markjay4k]

@robotdude17 With seq_len = 50, and normalise_window = True, both y_train and y_test will have all entries equal to -2.

I would suggest not normalizing the window, i.e. normalise_window = False

basically use this line in your code above instead

X_train, y_train, X_test, y_test = lstm.load_data('sinwave.csv', seq_len, False)