BUG: Unexpected KeyError in `transform()` with dict-like func argument

[sfo]

BUG: Unexpected KeyError in `transform()` with dict-like func argument

[Aloqeely]

You can pass axis="index" to apply the function to each column (I'm not quite sure what that means) per the doc at https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.transform.html

But I will admit this is a little confusing, I don't know why the error message says Column(s) [...] do not exist, the error message should be improved.
And I also find the docs to be a little inconsistent when it comes to the axis parameter, some docs say axis="index" will compute row-wise, and some say apply on each column, the only issue related issue I could found is #43128

[rhshadrach]

But I will admit this is a little confusing, I don't know why the error message says Column(s) [...] do not exist, the error message should be improved.

Agreed.

And I also find the docs to be a little inconsistent when it comes to the axis parameter, some docs say axis="index" will compute row-wise, and some say apply on each column, the only issue related issue I could found is #43128

Best to see the context, but I believe row-wise is meant as "down the rows", which is the same as applying to each column.

[luke396]

take