o # Old behavior o df.prod() # Would ignore non-numeric columns o # Future behavior o df.prod() # Raises TypeError