#1660 added numeric fields normalization for supp datasets#1676
Conversation
| return series.astype(float, errors="ignore").astype(str) | ||
|
|
||
| @staticmethod | ||
| def normalize_numeric_key(series: pd.Series) -> pd.Series: |
There was a problem hiding this comment.
In connected ticket @SFJohnson24 recommended moving this function in utils so both check operators and preprocessing logic can access the same function. Could you please update the code accordingly?
There was a problem hiding this comment.
should we keep both functions? since normalize_numeric_key applied to series and custom_str_conversion used per value? and use cases are a bit different...
There was a problem hiding this comment.
@SFJohnson24 what is your recommendation for this please?
There was a problem hiding this comment.
I still thinking moving custom_str_conversion to utils and utilizing that makes the most sense. Custom_str_conversion already handles the exact problem this feature describes--normalizing numeric string representations before comparison. I think we should harmonize the 2 as a single, reusable utils function instead of introducing a duplicate implementation.
There was a problem hiding this comment.
@SFJohnson24 The PR looks good to me now. Could you please confirm if your requested change is fully addressed. I will do a final validaiton and approve after that.
…ge to scalar function
SFJohnson24
left a comment
SFJohnson24
left a comment
There was a problem hiding this comment.
This PR correctly meets AC--moves string conversion to general util; does numeric check--merges on string key and restores original datatype to key column afterwards.
3 participants
No description provided.