Is there any clean way of setting numpy to use float32 values instead of float64 globally?

python numpy numbers
Edit
Report