Function _exact_ratio

tools/python-3.11.9-amd64/Lib/statistics.py:287–330

Return Real number x to exact (numerator, denominator) pair. >>> _exact_ratio(0.25) (1, 4) x is expected to be an int, Fraction, Decimal or float.

(x)

def _exact_ratio(x):
    """Return Real number x to exact (numerator, denominator) pair.

    >>> _exact_ratio(0.25)
    (1, 4)

    x is expected to be an int, Fraction, Decimal or float.
    """

    # XXX We should revisit whether using fractions to accumulate exact
    # ratios is the right way to go.

    # The integer ratios for binary floats can have numerators or
    # denominators with over 300 decimal digits.  The problem is more
    # acute with decimal floats where the default decimal context
    # supports a huge range of exponents from Emin=-999999 to
    # Emax=999999.  When expanded with as_integer_ratio(), numbers like
    # Decimal('3.14E+5000') and Decimal('3.14E-5000') have large
    # numerators or denominators that will slow computation.

    # When the integer ratios are accumulated as fractions, the size
    # grows to cover the full range from the smallest magnitude to the
    # largest.  For example, Fraction(3.14E+300) + Fraction(3.14E-300),
    # has a 616 digit numerator.  Likewise,
    # Fraction(Decimal('3.14E+5000')) + Fraction(Decimal('3.14E-5000'))
    # has 10,003 digit numerator.

    # This doesn't seem to have been problem in practice, but it is a
    # potential pitfall.

    try:
        return x.as_integer_ratio()
    except AttributeError:
        pass
    except (OverflowError, ValueError):
        # float NAN or INF.
        assert not _isfinite(x)
        return (x, None)
    try:
        # x may be an Integral ABC.
        return (x.numerator, x.denominator)
    except AttributeError:
        msg = f"can't convert type '{type(x).__name__}' to numerator/denominator"
        raise TypeError(msg)


def _convert(value, T):