You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Good day,
I was wondering, when using the .predict function in combination with timestamps (column 'date') there is no 'date' column in the prediction. But what is the resulting frequency of the prediction? I tested some models but never got a definitive answer.
Thanks
The text was updated successfully, but these errors were encountered:
By default, we use a sliding window of stride 1, and each prediction window moves forward one step at a time. For example, with a seq_len of 96, the first prediction is based on [0, 96), the second on [1, 97), and so on.
Hope this helps!
I'm afraid not.
My question is.
Inputing a timeseries with a column 'date' - where the timestamps are not equally spaced apart.
I get as predictions only timeseries without a 'date' column (so one column less than the input).
But what exactly are the timestamps/the time intervalls then?
Doing some tests i did not found a definitive answer, probably on my side though.
Good day,
I was wondering, when using the .predict function in combination with timestamps (column 'date') there is no 'date' column in the prediction. But what is the resulting frequency of the prediction? I tested some models but never got a definitive answer.
Thanks
The text was updated successfully, but these errors were encountered: