0
$\begingroup$

So the output vector I'm training a ktrain model on is 2 dimensional (the outputs are ['Temp', 'Velo']. And ktrain is even able to compute these when I use predictor.predict (see attached image), but when I store it in a variable, it returns a completely different 1 dimensional array of the same length. A screenshot of my code

Why is this happening? Where are those final 1 dimensional values even coming from? And how can I retrieve the actual 2 dimensional predicted output? Or will I necessarily have to train separate models for both vectors?

Improve this question
$\endgroup$
1
  • $\begingroup$ "UserWarning: is_regression=True was supplied but ignored because multiple label columns imply classification" is the warning I'm getting when I'm defining the model with the argument label_columns=['Temp', 'Velo']. Not sure that warning makes sense to me. $\endgroup$
    – Sisyphus
    Jun 1, 2023 at 20:23

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.