Making Multioutput Forecasts without Exogenous Data #4617
Unanswered
RahulDas-dev
asked this question in
Q&A
Replies: 1 comment 1 reply
-
I'm not an expert with Sktime, but I believe what is happening in your second example is you are including transformations on X in the forecasting pipeline. If you're trying to transform X in the train step, but not include X in the predict step, I think you'll need to do the transformation on X outside of the forecasting pipeline. |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hi All,
Thanks for the Beautiful Package, it is really great help for me.
My doubt is why
X [ Exogenous Data ]
is needed when we are doingmultioutput
forecasting.I am really stuck with this issue , Example 1 is working with out
X_data
but Example 2 is notExample 1
Example 2
Errors are given bellow
I would greatly appreciate any advice, suggestions, or alternative approaches from the community , Thanks in advance
Beta Was this translation helpful? Give feedback.
All reactions