Greetings everybody! I'm studying the neuralforecast cross-validation tutorial: https://nixtlaverse.nixtla.io/neuralforecast/docs/capabilities/cross_validation.html I changed the Y_df not to be filtered and consider all observations (h1...h99) commenting the following cell:

# Y_df = Y_df.query("unique_id == 'H1'")[:700]

# Y_df

Then i got the attached images. Does anyone knows what are those crossed lines?

It means you have overlapping windows, so you have multiple values for the same timestep. In the tutorial, we show an example of plotting with overlapping windows to avoid those crossed lines.

Thanks @Marco! Why doesn't the crossed lines appears when only H1 is analyzed? (ie, when the query filters

Y_df.query("unique_id == 'H1'")[:700]

). Aren't overlapping windows determined by the nf.cross_validation method parameters?

I'm not sure why, for the tutorial I only considered a single series, I don't know what the others look like. For sure, this is a plotting issue, and not an issue related to cross-validation

Ok, that's nice to know, at most some steps will be calculated more than once, right?

If

step_size

<

h

, yes