Fix `DataRow[column]` when column instance differs from row schema

### Problem

`DataRow[column]` behaves incorrectly when the provided `DataColumn` instance does not belong to the same schema as the row, even if a column with the same name exists.

This happens for example in `groupBy`, where:

- `df.x` (from the original DataFrame)
- `x` inside `groupBy`

represent different column instances.

As a result:

`key[df.x]` returns an incorrect value

instead of resolving the column by name.