Can't reproduce the results of standalone_simple_example.ipynb

[hangyao]

Hi,

I'm trying to reproduce the results in examples/keras/standalone_simple_example.ipynb, but the genrated data are different than what is shown in this notebook.

Below are the data plotted in my local notebook. I use Python 3.5 and Windows 7 OS. With these data sets, the neural network doesn't learn at all.

Another notebook simple_example.ipynb doesn't have this issue.

[tom-christie]

@hangyao I'm working through this notebook too. I think this line:

x_train = np.tile(x_train.T,n_repeats).T

should be:

x_train = np.repeat(x_train,n_repeats, axis=0)

There's a (unintentional?) time shift that makes the vertical lines in x_test become staggered in x_train. Changing the line in question fixes that problem and makes x_train look a lot more like x_test.

There's still some other problem I'm running into, as the learned alpha value is very high and constant (near 40). If I figure that one out I'll report back.

[ragulpr]

Thanks for reporting, I haven't run standalone_simple_example.ipynb for a while, it must be numpy-api changes or something of the sort. I'd copy paste use the data-generating code from simple_example.ipynb instead, that works.

Edit: Just saw this is over 1 year old, very sorry for missing this!