bit_truncate_float

caput.util.truncate.bit_truncate_float(val: numpy.float32, err: numpy.float32) → numpy.float32

Truncate using a fixed error.

Parameters:

valfloat32

The value to truncate.

errfloat32

The absolute precision to allow.

Returns:

truncatedfloat32

The truncated value.

Raises:

ValueError

If err is a NaN.