SNOW-977836: The withColumnRenamed fucntion fails to rename a column if the snowpark dataframe has multiple columns with same name but with different case style about snowpark-python HOT 2 OPEN

If by design, duplicate column names are allowable in Snowpark's API, then I think we should allow a way for the user to edit column names which are duplicated.

Example:

should be able to be changed to:

Instead of just throwing the error you showed.

from snowpark-python.

Hello @suenalaba, Sry If I was not clear. By definition we cannot create a snowpark dataframe with ambigious column names but we can create a snowpark dataframe that has two columns with same name but different case (i.e)
Name of column 1 - "Snow Flake"
Name of column 2 - "SNOW FLAKE"

Example dataframe :-

When we try to rename Column - 1 (i.e) "Snow Flake" to "Snow Flake Renamed" using the below command
df.with_column_renamed('"Snow Flake"','"Snow Flake Renamed"').show()

we run into an exception -> SnowparkColumnException: Unable to rename the column "Snow Flake" as "Snow Flake Renamed" because this DataFrame has 2 columns named "Snow Flake".

This is because, in the current implementation, the with_column_renamed method, the column being renamed is converted to upper case and then it is checked if there are any more columns with same name, In our case, we have a column that matches upper("Snow Flake") in snowpark dataframe and hence we run into this exception. I've addressed this in my PR #1149

from snowpark-python.