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
{{ message }}
This repository has been archived by the owner on Nov 29, 2024. It is now read-only.
Currently, if the input request contains columns that do not match the columns expected by the pipeline, these are silently ignored and the input has NaNs instead. This leads to rather cryptic behavior in real life cases where the input request contains a typo or a lower/upper case differences in the column names.
It would be safer to return error on these requests to make the discrepancy obvious and let the caller decide if an empty value should be used instead or not (= fail fast).
The text was updated successfully, but these errors were encountered:
Currently, if the input request contains columns that do not match the columns expected by the pipeline, these are silently ignored and the input has
NaN
s instead. This leads to rather cryptic behavior in real life cases where the input request contains a typo or a lower/upper case differences in the column names.It would be safer to return error on these requests to make the discrepancy obvious and let the caller decide if an empty value should be used instead or not (= fail fast).
The text was updated successfully, but these errors were encountered: