How does NumPy handle dtype casting rules internally?

NumPy uses a type hierarchy to safely cast values during mixed-type operations.

When operations involve multiple dtypes, NumPy applies a casting hierarchy to determine a safe resulting type. This prevents data loss where possible. For example, int + float results in float. The system ensures consistency across ufuncs, reductions, and assignments. Unsafe casts are either disallowed or require explicit conversion.