Tensorflow Tutorial on Time Series Forecasting

TensorFlow General Discussion
1

Hello, everyone!

I am following Tensorflow’s tutorial on time series forecasting (Time series forecasting  |  TensorFlow Core), and I can’t understand the graphics generated by the plot() function defined in topic 3:

def plot(self, model=None, plot_col='T (degC)', max_subplots=3):  inputs, labels = self.example
``` .....

How can we determine what samples are being plotted here? How could I retrieve datetimes related with each of them?

Any help will be appreciated.

Kind regards,
3

Hi,
Date and time columns are given in the dataset. If you check initial dataset before pop method, you can see them. Function to_datetime() converts them to datetime object and then its used as an index variable.

image
image1875×794 53.9 KB

Regarding which samples are plotted, in plot function , the default argument is 'T (degC) so first call ```
w2.plot() doesnt have any argument and it will plot temperature. Second time the call is w2.plot(plot_col=‘p (mbar)’) so it will plot pressure and so on.
I hope its clear to you.