passagemath-categories 10.6.31rc3__cp314-cp314-musllinux_1_2_aarch64.whl

sage/rings/real_double.pyx

@@ -0,0 +1,2215 @@
+# sage_setup: distribution = sagemath-categories
+r"""
+Double precision floating point real numbers
+
+EXAMPLES:
+
+We create the real double vector space of dimension `3`::
+
+    sage: V = RDF^3; V                                                                  # needs sage.modules
+    Vector space of dimension 3 over Real Double Field
+
+Notice that this space is unique::
+
+    sage: V is RDF^3                                                                    # needs sage.modules
+    True
+    sage: V is FreeModule(RDF, 3)                                                       # needs sage.modules
+    True
+    sage: V is VectorSpace(RDF, 3)                                                      # needs sage.modules
+    True
+
+Also, you can instantly create a space of large dimension::
+
+    sage: V = RDF^10000                                                                 # needs sage.modules
+
+TESTS:
+
+Test NumPy conversions::
+
+    sage: RDF(1).__array_interface__
+    {'typestr': '=f8'}
+    sage: import numpy                                                                  # needs numpy
+    sage: numpy.array([RDF.pi()]).dtype                                                 # needs numpy
+    dtype('float64')
+"""
+
+# ****************************************************************************
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 2 of the License, or
+# (at your option) any later version.
+#                  https://www.gnu.org/licenses/
+# ****************************************************************************
+
+cimport libc.math
+from libc.string cimport memcpy
+from cpython.object cimport *
+from cpython.float cimport *
+
+from sage.ext.stdsage cimport PY_NEW
+from sage.cpython.python_debug cimport if_Py_TRACE_REFS_then_PyObject_INIT
+
+import math
+
+import sage.arith.misc
+import sage.rings.integer
+import sage.rings.rational
+
+from sage.rings.integer cimport Integer
+from sage.rings.integer_ring import ZZ
+
+from sage.categories.morphism cimport Morphism
+from sage.structure.coerce cimport is_numpy_type
+from sage.misc.randstate cimport randstate, current_randstate
+from sage.structure.richcmp cimport rich_to_bool
+from sage.arith.constants cimport *
+
+cimport gmpy2
+
+
+new_gen_from_real_double_element = None
+
+
+cdef class RealDoubleField_class(sage.rings.abc.RealDoubleField):
+    """
+    An approximation to the field of real numbers using double
+    precision floating point numbers. Answers derived from calculations
+    in this approximation may differ from what they would be if those
+    calculations were performed in the true field of real numbers. This
+    is due to the rounding errors inherent to finite precision
+    calculations.
+
+    EXAMPLES::
+
+        sage: RR == RDF                                                                 # needs sage.rings.real_mpfr
+        False
+        sage: RDF == RealDoubleField()    # RDF is the shorthand
+        True
+
+    ::
+
+        sage: RDF(1)
+        1.0
+        sage: RDF(2/3)
+        0.6666666666666666
+
+    A :exc:`TypeError` is raised if the coercion doesn't make sense::
+
+        sage: RDF(QQ['x'].0)
+        Traceback (most recent call last):
+        ...
+        TypeError: cannot convert nonconstant polynomial
+        sage: RDF(QQ['x'](3))
+        3.0
+
+    One can convert back and forth between double precision real
+    numbers and higher-precision ones, though of course there may be
+    loss of precision::
+
+        sage: # needs sage.rings.real_mpfr
+        sage: a = RealField(200)(2).sqrt(); a
+        1.4142135623730950488016887242096980785696718753769480731767
+        sage: b = RDF(a); b
+        1.4142135623730951
+        sage: a.parent()(b)
+        1.4142135623730951454746218587388284504413604736328125000000
+        sage: a.parent()(b) == b
+        True
+        sage: b == RR(a)
+        True
+
+    TESTS::
+
+        sage: RDF.is_finite()
+        False
+    """
+    def __init__(self):
+        """
+        Initialize ``self``.
+
+        TESTS::
+
+            sage: R = RealDoubleField()
+            sage: TestSuite(R).run()
+        """
+        from sage.categories.fields import Fields
+        Field.__init__(self, self,
+                       category=Fields().Infinite().Metric().Complete())
+        self._populate_coercion_lists_(init_no_parent=True,
+                                       convert_method_name='_real_double_')
+
+    _element_constructor_ = RealDoubleElement
+
+    def __reduce__(self):
+        """
+        For pickling.
+
+        EXAMPLES::
+
+            sage: loads(dumps(RDF)) is RDF
+            True
+        """
+        return RealDoubleField, ()
+
+    cpdef bint is_exact(self) except -2:
+        """
+        Return ``False``, because doubles are not exact.
+
+        EXAMPLES::
+
+            sage: RDF.is_exact()
+            False
+        """
+        return False
+
+    def _latex_(self):
+        r"""
+        Return a latex representation of ``self``.
+
+        EXAMPLES::
+
+            sage: latex(RDF)  # indirect doctest
+            \Bold{R}
+        """
+        return "\\Bold{R}"
+
+    def _sage_input_(self, sib, coerced):
+        r"""
+        Produce an expression which will reproduce this value when evaluated.
+
+        EXAMPLES::
+
+            sage: sage_input(RDF, verify=True)
+            # Verified
+            RDF
+            sage: from sage.misc.sage_input import SageInputBuilder
+            sage: RDF._sage_input_(SageInputBuilder(), False)
+            {atomic:RDF}
+        """
+        return sib.name('RDF')
+
+    def __repr__(self):
+        """
+        Return a string representation of ``self``.
+
+        EXAMPLES::
+
+            sage: RealDoubleField()  # indirect doctest
+            Real Double Field
+            sage: RDF
+            Real Double Field
+        """
+        return "Real Double Field"
+
+    def _repr_option(self, key):
+        """
+        Metadata about the :meth:`_repr_` output.
+
+        See :meth:`sage.structure.parent._repr_option` for details.
+
+        EXAMPLES::
+
+            sage: RDF._repr_option('element_is_atomic')
+            True
+        """
+        if key == 'element_is_atomic':
+            return True
+        return super()._repr_option(key)
+
+    def __richcmp__(self, x, op):
+        """
+        Compare ``self`` to ``x``.
+
+        EXAMPLES::
+
+            sage: RDF == 5
+            False
+            sage: loads(dumps(RDF)) == RDF
+            True
+        """
+        if isinstance(x, RealDoubleField_class):
+            return rich_to_bool(op, 0)
+        if op == Py_NE:
+            return True
+        return NotImplemented
+
+    def construction(self):
+        r"""
+        Return the functorial construction of ``self``, namely, completion of
+        the rational numbers with respect to the prime at `\infty`.
+
+        Also preserves other information that makes this field unique (i.e.
+        the Real Double Field).
+
+        EXAMPLES::
+
+            sage: c, S = RDF.construction(); S
+            Rational Field
+            sage: RDF == c(S)
+            True
+        """
+        from sage.categories.pushout import CompletionFunctor
+        return (CompletionFunctor(sage.rings.infinity.Infinity,
+                                  53,
+                                  {'type': 'RDF'}),
+                sage.rings.rational_field.QQ)
+
+    def complex_field(self):
+        """
+        Return the complex field with the same precision as ``self``, i.e.,
+        the complex double field.
+
+        EXAMPLES::
+
+            sage: RDF.complex_field()                                                   # needs sage.rings.complex_double
+            Complex Double Field
+        """
+        from sage.rings.complex_double import CDF
+        return CDF
+
+    def algebraic_closure(self):
+        """
+        Return the algebraic closure of ``self``, i.e., the complex double
+        field.
+
+        EXAMPLES::
+
+            sage: RDF.algebraic_closure()                                               # needs sage.rings.complex_double
+            Complex Double Field
+        """
+        from sage.rings.complex_double import CDF
+        return CDF
+
+    cpdef _coerce_map_from_(self, S):
+        """
+        Canonical coercion of ``S`` to the real double field.
+
+        The rings that canonically coerce to the real double field are:
+
+        - the real double field itself
+        - int, long, integer, and rational rings
+        - numpy integers and floatings
+        - the real lazy field
+        - the MPFR real field with at least 53 bits of precision
+
+        EXAMPLES::
+
+            sage: RDF.coerce(5)  # indirect doctest
+            5.0
+            sage: RDF.coerce(9499294r)
+            9499294.0
+            sage: RDF.coerce(61/3)
+            20.333333333333332
+            sage: parent(RDF(3) + CDF(5))                                               # needs sage.rings.complex_double
+            Complex Double Field
+            sage: parent(CDF(5) + RDF(3))                                               # needs sage.rings.complex_double
+            Complex Double Field
+            sage: CDF.gen(0) + 5.0                                                      # needs sage.rings.complex_double
+            5.0 + 1.0*I
+            sage: RLF(2/3) + RDF(1)
+            1.6666666666666665
+
+            sage: import numpy                                                          # needs numpy
+            sage: RDF.coerce(numpy.int8('1'))                                           # needs numpy
+            1.0
+            sage: RDF.coerce(numpy.float64('1'))                                        # needs numpy
+            1.0
+
+            sage: RDF.coerce(pi)                                                        # needs sage.symbolic
+            Traceback (most recent call last):
+            ...
+            TypeError: no canonical coercion from Symbolic Ring to Real Double Field
+
+        Test that :issue:`15695` is fixed (see also :issue:`18076`)::
+
+            sage: 1j + numpy.float64(2)                                                 # needs numpy
+            2.00000000000000 + 1.00000000000000*I
+            sage: parent(_)                                                             # needs numpy
+            Complex Field with 53 bits of precision
+        """
+        if S is int or S is float:
+            return ToRDF(S)
+
+        from sage.rings.rational_field import QQ
+        try:
+            from sage.rings.real_lazy import RLF
+        except ImportError:
+            RLF = None
+
+        if S is ZZ or S is QQ or S is RLF:
+            return ToRDF(S)
+
+        if isinstance(S, sage.rings.abc.RealField):
+            if S.prec() >= 53:
+                return ToRDF(S)
+            else:
+                return None
+        elif is_numpy_type(S):
+            import numpy
+            if issubclass(S, numpy.integer) or issubclass(S, numpy.floating):
+                return ToRDF(S)
+            else:
+                return None
+
+        try:
+            from sage.rings.real_mpfr import RR
+        except ImportError:
+            pass
+        else:
+            connecting = RR._internal_coerce_map_from(S)
+            if connecting is not None:
+                return ToRDF(RR) * connecting
+
+    def _magma_init_(self, magma):
+        r"""
+        Return a string representation of ``self`` in the Magma language.
+
+        EXAMPLES:
+
+        Magma handles precision in decimal digits, so we lose a bit::
+
+            sage: magma(RDF)        # indirect doctest  # optional - magma
+            Real field of precision 15
+            sage: 10^15 < 2^53 < 10^16
+            True
+
+        When we convert back from Magma, we convert to a generic real field
+        that has 53 bits of precision::
+
+            sage: magma(RDF).sage()                     # optional - magma
+            Real Field with 53 bits of precision
+        """
+        return "RealField(%s : Bits := true)" % self.prec()
+
+    def _fricas_init_(self):
+        r"""
+        Return the FriCAS representation of the real double field.
+
+        EXAMPLES::
+
+            sage: fricas(RDF)       # indirect doctest  # optional - fricas
+            DoubleFloat
+        """
+        return "DoubleFloat"
+
+    def _polymake_init_(self):
+        r"""
+        Return the polymake representation of the real double field.
+
+        EXAMPLES::
+
+            sage: polymake(RDF)     # indirect doctest  # optional - jupymake
+            Float
+        """
+        return '"Float"'
+
+    def precision(self):
+        """
+        Return the precision of this real double field in bits.
+
+        Always returns 53.
+
+        EXAMPLES::
+
+            sage: RDF.precision()
+            53
+        """
+        return 53
+
+    prec = precision
+
+    def to_prec(self, prec):
+        """
+        Return the real field to the specified precision. As doubles have
+        fixed precision, this will only return a real double field if ``prec``
+        is exactly 53.
+
+        EXAMPLES::
+
+            sage: RDF.to_prec(52)                                                       # needs sage.rings.real_mpfr
+            Real Field with 52 bits of precision
+            sage: RDF.to_prec(53)
+            Real Double Field
+        """
+        if prec == 53:
+            return self
+        from sage.rings.real_mpfr import RealField
+        return RealField(prec)
+
+    def gen(self, n=0):
+        """
+        Return the generator of the real double field.
+
+        EXAMPLES::
+
+            sage: RDF.0
+            1.0
+            sage: RDF.gens()
+            (1.0,)
+        """
+        if n != 0:
+            raise ValueError("only 1 generator")
+        return RealDoubleElement(1)
+
+    def ngens(self):
+        """
+        Return the number of generators which is always 1.
+
+        EXAMPLES::
+
+            sage: RDF.ngens()
+            1
+        """
+        return 1
+
+    def characteristic(self):
+        """
+        Return 0, since the field of real numbers has characteristic 0.
+
+        EXAMPLES::
+
+            sage: RDF.characteristic()
+            0
+        """
+        return Integer(0)
+
+    cdef _new_c(self, double value):
+        cdef RealDoubleElement x
+        x = PY_NEW(RealDoubleElement)
+        x._value = value
+        return x
+
+    def random_element(self, double min=-1, double max=1):
+        """
+        Return a random element of this real double field in the interval
+        ``[min, max]``.
+
+        EXAMPLES::
+
+            sage: RDF.random_element().parent() is RDF
+            True
+            sage: -1 <= RDF.random_element() <= 1
+            True
+            sage: 100 <= RDF.random_element(min=100, max=110) <= 110
+            True
+        """
+        cdef randstate rstate = current_randstate()
+
+        return self._new_c((max-min)*rstate.c_rand_double() + min)
+
+    def name(self):
+        """
+        The name of ``self``.
+
+        EXAMPLES::
+
+            sage: RDF.name()
+            'RealDoubleField'
+        """
+        return "RealDoubleField"
+
+    def __hash__(self):
+        """
+        Return the hash value of ``self``.
+
+        This class is intended for use as a singleton so any instance
+        of it should be equivalent from a hashing perspective.
+
+        TESTS::
+
+            sage: from sage.rings.real_double import RealDoubleField_class
+            sage: hash(RDF) == hash(RealDoubleField_class())
+            True
+        """
+        return 1157042230
+
+    def pi(self):
+        r"""
+        Return `\pi` to double-precision.
+
+        EXAMPLES::
+
+            sage: RDF.pi()
+            3.141592653589793
+            sage: RDF.pi().sqrt()/2
+            0.8862269254527579
+        """
+        return self(M_PI)
+
+    def euler_constant(self):
+        """
+        Return Euler's gamma constant to double precision.
+
+        EXAMPLES::
+
+            sage: RDF.euler_constant()
+            0.5772156649015329
+        """
+        return self(M_EULER)
+
+    def log2(self):
+        r"""
+        Return `\log(2)` to the precision of this field.
+
+        EXAMPLES::
+
+            sage: RDF.log2()
+            0.6931471805599453
+            sage: RDF(2).log()
+            0.6931471805599453
+        """
+        return self(M_LN2)
+
+    def factorial(self, int n):
+        """
+        Return the factorial of the integer `n` as a real number.
+
+        EXAMPLES::
+
+            sage: RDF.factorial(100)
+            9.332621544394415e+157
+        """
+        return global_dummy_element._factorial(n)
+
+    def zeta(self, n=2):
+        """
+        Return an `n`-th root of unity in the real field, if one
+        exists, or raise a :exc:`ValueError` otherwise.
+
+        EXAMPLES::
+
+            sage: RDF.zeta()
+            -1.0
+            sage: RDF.zeta(1)
+            1.0
+            sage: RDF.zeta(5)
+            Traceback (most recent call last):
+            ...
+            ValueError: No 5th root of unity in self
+        """
+        if n == 1:
+            return self(1)
+        elif n == 2:
+            return self(-1)
+        raise ValueError("No %sth root of unity in self" % n)
+
+    def NaN(self):
+        """
+        Return Not-a-Number ``NaN``.
+
+        EXAMPLES::
+
+            sage: RDF.NaN()
+            NaN
+        """
+        return self(0)/self(0)
+
+    nan = NaN
+
+    def _factor_univariate_polynomial(self, f):
+        """
+        Factor the univariate polynomial ``f``.
+
+        INPUT:
+
+        - ``f`` -- a univariate polynomial defined over the double precision
+          real numbers
+
+        OUTPUT:
+
+        - A factorization of ``f`` over the double precision real numbers
+          into a unit and monic irreducible factors
+
+        .. NOTE::
+
+            This is a helper method for
+            :meth:`sage.rings.polynomial.polynomial_element.Polynomial.factor`.
+
+        TESTS::
+
+            sage: # needs numpy
+            sage: R.<x> = RDF[]
+            sage: RDF._factor_univariate_polynomial(x)
+            x
+            sage: RDF._factor_univariate_polynomial(2*x)
+            (2.0) * x
+            sage: RDF._factor_univariate_polynomial(x^2)
+            x^2
+            sage: RDF._factor_univariate_polynomial(x^2 + 1)
+            x^2 + 1.0
+            sage: RDF._factor_univariate_polynomial(x^2 - 1)
+            (x - 1.0) * (x + 1.0)
+
+        The implementation relies on the ``roots()`` method which often reports
+        roots not to be real even though they are::
+
+            sage: f = (x-1)^3                                                           # needs numpy
+            sage: f.roots(ring=CDF)  # abs tol 2e-5                                     # needs numpy
+            [(1.0000065719436413, 1),
+             (0.9999967140281792 - 5.691454546815028e-06*I, 1),
+             (0.9999967140281792 + 5.691454546815028e-06*I, 1)]
+
+        This leads to the following incorrect factorization::
+
+            sage: f.factor()  # abs tol 2e-5                                            # needs numpy
+            (x - 1.0000065719436413) * (x^2 - 1.9999934280563585*x + 0.9999934280995487)
+        """
+        from sage.rings.complex_double import CDF
+        roots = f.roots(CDF)
+
+        # collect real roots and conjugate pairs of non-real roots
+        real_roots = [(r, e) for r, e in roots if r.imag().is_zero()]
+        non_real_roots = {r: e for r, e in roots if not r.imag().is_zero()}
+        assert all(non_real_roots[r.conj()] == e for r, e in non_real_roots.items()), "Bug in root finding code over RDF - roots must always come in conjugate pairs"
+        non_real_roots = [(r, e) for r, e in non_real_roots.items() if r.imag() > 0]
+
+        # turn the roots into irreducible factors
+        x = f.parent().gen()
+        real_factors = [(x - r.real(), e) for r, e in real_roots]
+        non_real_factors = [(x**2 - (r + r.conj()).real()*x + (r*r.conj()).real(), e) for r, e in non_real_roots]
+
+        # make the factors monic
+        from sage.structure.factorization import Factorization
+        return Factorization([(g.monic(), e) for g, e in real_factors + non_real_factors], f.leading_coefficient())
+
+
+cdef class RealDoubleElement(FieldElement):
+    """
+    An approximation to a real number using double precision floating
+    point numbers. Answers derived from calculations with such
+    approximations may differ from what they would be if those
+    calculations were performed with true real numbers. This is due to
+    the rounding errors inherent to finite precision calculations.
+    """
+
+    __array_interface__ = {'typestr': '=f8'}
+
+    def __cinit__(self):
+        """
+        Initialize ``self`` for cython.
+
+        EXAMPLES::
+
+            sage: RDF(2.3)  # indirect doctest
+            2.3
+        """
+        (<Element>self)._parent = _RDF
+
+    def __init__(self, x):
+        """
+        Create a new ``RealDoubleElement`` with value ``x``.
+
+        EXAMPLES::
+
+            sage: RDF(10^100)
+            1e+100
+
+        TESTS::
+
+            sage: from gmpy2 import *
+            sage: RDF(mpz(42))
+            42.0
+            sage: RDF(mpq(3/4))
+            0.75
+            sage: RDF(mpq('4.1'))
+            4.1
+        """
+        self._value = float(x)
+
+    def _magma_init_(self, magma):
+        r"""
+        Return a string representation of ``self`` in the Magma language.
+
+        EXAMPLES::
+
+            sage: RDF(10.5)
+            10.5
+            sage: magma(RDF(10.5))  # indirect doctest  # optional - magma
+            10.5000000000000
+        """
+        return "%s!%s" % (self.parent()._magma_init_(magma), self)
+
+    def __reduce__(self):
+        """
+        For pickling.
+
+        EXAMPLES::
+
+            sage: a = RDF(-2.7)
+            sage: loads(dumps(a)) == a
+            True
+        """
+        return RealDoubleElement, (self._value, )
+
+    cdef _new_c(self, double value):
+        cdef RealDoubleElement x
+        x = PY_NEW(RealDoubleElement)
+        x._value = value
+        return x
+
+    def prec(self):
+        """
+        Return the precision of this number in bits.
+
+        Always returns 53.
+
+        EXAMPLES::
+
+            sage: RDF(0).prec()
+            53
+        """
+        return 53
+
+    def ulp(self):
+        """
+        Return the unit of least precision of ``self``, which is the
+        weight of the least significant bit of ``self``. This is always
+        a strictly positive number. It is also the gap between this
+        number and the closest number with larger absolute value that
+        can be represented.
+
+        EXAMPLES::
+
+            sage: a = RDF(pi)                                                           # needs sage.symbolic
+            sage: a.ulp()                                                               # needs sage.symbolic
+            4.440892098500626e-16
+            sage: b = a + a.ulp()                                                       # needs sage.symbolic
+
+        Adding or subtracting an ulp always gives a different number::
+
+            sage: # needs sage.symbolic
+            sage: a + a.ulp() == a
+            False
+            sage: a - a.ulp() == a
+            False
+            sage: b + b.ulp() == b
+            False
+            sage: b - b.ulp() == b
+            False
+
+        Since the default rounding mode is round-to-nearest, adding or
+        subtracting something less than half an ulp always gives the
+        same number, unless the result has a smaller ulp. The latter
+        can only happen if the input number is (up to sign) exactly a
+        power of 2::
+
+            sage: # needs sage.symbolic
+            sage: a - a.ulp()/3 == a
+            True
+            sage: a + a.ulp()/3 == a
+            True
+            sage: b - b.ulp()/3 == b
+            True
+            sage: b + b.ulp()/3 == b
+            True
+
+            sage: c = RDF(1)
+            sage: c - c.ulp()/3 == c
+            False
+            sage: c.ulp()
+            2.220446049250313e-16
+            sage: (c - c.ulp()).ulp()
+            1.1102230246251565e-16
+
+        The ulp is always positive::
+
+            sage: RDF(-1).ulp()
+            2.220446049250313e-16
+
+        The ulp of zero is the smallest positive number in RDF::
+
+            sage: RDF(0).ulp()
+            5e-324
+            sage: RDF(0).ulp()/2
+            0.0
+
+        Some special values::
+
+            sage: a = RDF(1)/RDF(0); a
+            +infinity
+            sage: a.ulp()
+            +infinity
+            sage: (-a).ulp()
+            +infinity
+            sage: a = RDF('nan')
+            sage: a.ulp() is a
+            True
+
+        The ulp method works correctly with small numbers::
+
+            sage: u = RDF(0).ulp()
+            sage: u.ulp() == u
+            True
+            sage: x = u * (2^52-1)  # largest denormal number
+            sage: x.ulp() == u
+            True
+            sage: x = u * 2^52  # smallest normal number
+            sage: x.ulp() == u
+            True
+        """
+        # First, check special values
+        if self._value == 0:
+            return RealDoubleElement(libc.math.ldexp(1.0, -1074))
+        if libc.math.isnan(self._value):
+            return self
+        if libc.math.isinf(self._value):
+            return self.abs()
+
+        # Normal case
+        cdef int e
+        libc.math.frexp(self._value, &e)
+        e -= 53
+        # Correction for denormals
+        if e < -1074:
+            e = -1074
+        return RealDoubleElement(libc.math.ldexp(1.0, e))
+
+    def real(self):
+        """
+        Return ``self`` - we are already real.
+
+        EXAMPLES::
+
+            sage: a = RDF(3)
+            sage: a.real()
+            3.0
+        """
+        return self
+
+    def imag(self):
+        """
+        Return the imaginary part of this number, which is zero.
+
+        EXAMPLES::
+
+            sage: a = RDF(3)
+            sage: a.imag()
+            0.0
+        """
+        return RealDoubleElement(0)
+
+    def __complex__(self):
+        """
+        Return ``self`` as a python complex number.
+
+        EXAMPLES::
+
+            sage: a = 2303
+            sage: RDF(a)
+            2303.0
+            sage: complex(RDF(a))
+            (2303+0j)
+        """
+        return complex(self._value, 0)
+
+    def _integer_(self, ZZ=None):
+        """
+        If this floating-point number is actually an integer, return
+        that integer.  Otherwise, raise an exception.
+
+        EXAMPLES::
+
+            sage: ZZ(RDF(237.0))  # indirect doctest
+            237
+            sage: ZZ(RDF(0.0/0.0))
+            Traceback (most recent call last):
+            ...
+            ValueError: cannot convert float NaN to integer
+            sage: ZZ(RDF(1.0/0.0))
+            Traceback (most recent call last):
+            ...
+            OverflowError: cannot convert float infinity to integer
+            sage: ZZ(RDF(-123456789.0))
+            -123456789
+            sage: ZZ(RDF((2.0))^290)
+            1989292945639146568621528992587283360401824603189390869761855907572637988050133502132224
+            sage: ZZ(RDF(-2345.67))
+            Traceback (most recent call last):
+            ...
+            TypeError: cannot convert non-integral float to integer
+        """
+        return Integer(self._value)
+
+    def __mpfr__(self):
+        """
+        Convert Sage ``RealDoubleElement`` to gmpy2 ``mpfr``.
+
+        EXAMPLES::
+
+            sage: RDF(42.2).__mpfr__()
+            mpfr('42.200000000000003')
+            sage: from gmpy2 import mpfr
+            sage: mpfr(RDF(5.1))
+            mpfr('5.0999999999999996')
+
+        TESTS::
+
+            sage: RDF().__mpfr__(); raise NotImplementedError("gmpy2 is not installed")
+            Traceback (most recent call last):
+            ...
+            NotImplementedError: gmpy2 is not installed
+        """
+        return gmpy2.mpfr(self._value)
+
+    def _interface_init_(self, I=None):
+        """
+        Return ``self`` formatted as a string, suitable as input to another
+        computer algebra system. (This is the default function used for
+        exporting to other computer algebra systems.)
+
+        EXAMPLES::
+
+            sage: s1 = RDF(sin(1)); s1                                                  # needs sage.symbolic
+            0.8414709848078965
+            sage: s1._interface_init_()                                                 # needs sage.symbolic
+            '0.8414709848078965'
+            sage: s1 == RDF(gp(s1))                                                     # needs sage.libs.pari sage.symbolic
+            True
+        """
+        return repr(self._value)
+
+    def _mathematica_init_(self):
+        """
+        TESTS:
+
+        Check that :issue:`28814` is fixed::
+
+            sage: mathematica(RDF(1e25))   # optional - mathematica
+            1.*^25
+            sage: mathematica(RDF(1e-25))  # optional - mathematica
+            1.*^-25
+        """
+        from sage.rings.real_mpfr import RR
+        return RR(self._value)._mathematica_init_()
+
+    def _sage_input_(self, sib, coerced):
+        r"""
+        Produce an expression which will reproduce this value when evaluated.
+
+        EXAMPLES::
+
+            sage: sage_input(RDF(NaN))                                                  # needs sage.symbolic
+            RDF(NaN)
+            sage: sage_input(RDF(-infinity), verify=True)
+            # Verified
+            -RDF(infinity)
+            sage: sage_input(RDF(-infinity)*polygen(RDF))
+            R.<x> = RDF[]
+            -RDF(infinity)*x + RDF(NaN)
+            sage: sage_input(RDF(pi), verify=True)                                      # needs sage.symbolic
+            # Verified
+            RDF(3.1415926535897931)
+            sage: sage_input(RDF(-e), verify=True, preparse=False)                      # needs sage.symbolic
+            # Verified
+            -RDF(2.718281828459045...)
+            sage: sage_input(RDF(pi)*polygen(RDF), verify=True, preparse=None)          # needs sage.symbolic
+            # Verified
+            R = RDF['x']
+            x = R.gen()
+            3.1415926535897931*x
+            sage: from sage.misc.sage_input import SageInputBuilder
+            sage: sib = SageInputBuilder()
+            sage: RDF(22/7)._sage_input_(sib, True)                                     # needs sage.sage.rings.real_mpfr
+            {atomic:3.1428571428571428}
+            sage: RDF(22/7)._sage_input_(sib, False)                                    # needs sage.sage.rings.real_mpfr
+            {call: {atomic:RDF}({atomic:3.1428571428571428})}
+        """
+        cdef bint isinf = libc.math.isinf(self._value)
+        cdef bint isnan = libc.math.isnan(self._value)
+        if isinf or isnan:
+            if isnan:
+                v = sib.name('NaN')
+            else:
+                v = sib.name('infinity')
+            v = sib(self.parent())(v)
+            if self._value < 0:
+                v = -v
+            return v
+
+        from sage.rings.integer_ring import ZZ
+        from sage.rings.real_mpfr import RR
+
+        cdef bint negative = self._value < 0
+        if negative:
+            self = -self
+
+        # There are five possibilities for printing this floating-point
+        # number, ordered from prettiest to ugliest (IMHO).
+        # 1) An integer: 42
+        # 2) A simple literal: 3.14159
+        # 3) A coerced integer: RDF(42)
+        # 4) A coerced literal: RDF(3.14159)
+        # 5) A coerced RR value: RDF(RR('3.14159'))
+
+        # str(self) works via libc, which we don't necessarily trust
+        # to produce the best possible answer.  So this function prints
+        # via RR/MPFR.  Without the preparser, input works via libc as
+        # well, but we don't have a choice about that.
+
+        # To use choice 1 or choice 3, this number must be an integer.
+        cdef bint can_use_int_literal = \
+            self.abs() < (Integer(1) << self.prec()) and self in ZZ
+
+        self_str = RR(self._value).str(truncate=False, skip_zeroes=True)
+
+        # To use choice 2 or choice 4, we must be able to read
+        # numbers of this precision as a literal.
+        cdef bint can_use_float_literal = \
+            (sib.preparse() or float(self_str) == self)
+
+        if can_use_int_literal or can_use_float_literal:
+            if can_use_int_literal:
+                v = sib.int(self._integer_())
+            else:
+                v = sib.float_str(self_str)
+        else:
+            v = sib(RR(self))
+        if not coerced:
+            v = sib(self.parent())(v)
+
+        if negative:
+            v = -v
+
+        return v
+
+    def _repr_(self):
+        """
+        Return the string representation of ``self``.
+
+        EXAMPLES::
+
+            sage: RDF(-2/3)
+            -0.6666666666666666
+            sage: a = RDF(2); a
+            2.0
+            sage: a^2
+            4.0
+            sage: RDF("nan")
+            NaN
+            sage: RDF(1)/RDF(0)
+            +infinity
+            sage: RDF(-1)/RDF(0)
+            -infinity
+        """
+        return double_repr(self._value)
+
+    def __format__(self, format_spec):
+        """
+        Return a formatted string representation of this real number.
+
+        EXAMPLES::
+
+            sage: format(RDF(32/3), '.4f')
+            '10.6667'
+            sage: '{:.4e}'.format(RDF(2/3))
+            '6.6667e-01'
+        """
+        return format(float(self), format_spec)
+
+    def _latex_(self):
+        r"""
+        Return a latex representation of ``self``.
+
+        EXAMPLES::
+
+            sage: latex(RDF(3.4))       # indirect doctest
+            3.4
+            sage: latex(RDF(2e-100))    # indirect doctest
+            2 \times 10^{-100}
+        """
+        s = self.str()
+        parts = s.split('e')
+        if len(parts) > 1:
+            # scientific notation
+            if parts[1][0] == '+':
+                parts[1] = parts[1][1:]
+            s = "%s \\times 10^{%s}" % (parts[0], parts[1])
+        return s
+
+    def __hash__(self):
+        """
+        Return the hash of ``self``, which coincides with the python float
+        (and often int) type.
+
+        EXAMPLES::
+
+            sage: hash(RDF(1.2)) == hash(1.2r)
+            True
+        """
+        return hash(self._value)
+
+    def _im_gens_(self, codomain, im_gens, base_map=None):
+        """
+        Return the image of ``self`` under the homomorphism from the rational
+        field to ``codomain``.
+
+        This always just returns ``self`` coerced into the ``codomain``.
+
+        EXAMPLES::
+
+            sage: RDF(2.1)._im_gens_(RR, [RR(1)])
+            2.10000000000000
+            sage: R = RealField(20)                                                     # needs sage.rings.real_mpfr
+            sage: RDF(2.1)._im_gens_(R, [R(1)])                                         # needs sage.rings.real_mpfr
+            2.1000
+        """
+        return codomain(self)  # since 1 |--> 1
+
+    def str(self):
+        """
+        Return the informal string representation of ``self``.
+
+        EXAMPLES::
+
+            sage: a = RDF('4.5'); a.str()
+            '4.5'
+            sage: a = RDF('49203480923840.2923904823048'); a.str()
+            '49203480923840.29'
+            sage: a = RDF(1)/RDF(0); a.str()
+            '+infinity'
+            sage: a = -RDF(1)/RDF(0); a.str()
+            '-infinity'
+            sage: a = RDF(0)/RDF(0); a.str()
+            'NaN'
+
+        We verify consistency with ``RR`` (mpfr reals)::
+
+            sage: str(RR(RDF(1)/RDF(0))) == str(RDF(1)/RDF(0))
+            True
+            sage: str(RR(-RDF(1)/RDF(0))) == str(-RDF(1)/RDF(0))
+            True
+            sage: str(RR(RDF(0)/RDF(0))) == str(RDF(0)/RDF(0))
+            True
+        """
+        return double_repr(self._value)
+
+    def __copy__(self):
+        """
+        Return copy of ``self``, which since ``self`` is immutable, is just
+        ``self``.
+
+        EXAMPLES::
+
+            sage: r = RDF('-1.6')
+            sage: r.__copy__() is r
+            True
+        """
+        return self
+
+    def __deepcopy__(self, memo):
+        """
+        EXAMPLES::
+
+            sage: r = RDF('-1.6')
+            sage: deepcopy(r) is r
+            True
+        """
+        return self
+
+    def integer_part(self):
+        """
+        If in decimal this number is written ``n.defg``, returns ``n``.
+
+        EXAMPLES::
+
+            sage: r = RDF('-1.6')
+            sage: a = r.integer_part(); a
+            -1
+            sage: type(a)
+            <class 'sage.rings.integer.Integer'>
+            sage: r = RDF(0.0/0.0)
+            sage: a = r.integer_part()
+            Traceback (most recent call last):
+            ...
+            TypeError: Attempt to get integer part of NaN
+        """
+        if libc.math.isnan(self._value):
+            raise TypeError("Attempt to get integer part of NaN")
+        else:
+            return Integer(int(self._value))
+
+    def sign_mantissa_exponent(self):
+        r"""
+        Return the sign, mantissa, and exponent of ``self``.
+
+        In Sage (as in MPFR), floating-point numbers of precision `p`
+        are of the form `s m 2^{e-p}`, where `s \in \{-1, 1\}`,
+        `2^{p-1} \leq m < 2^p`, and `-2^{30} + 1 \leq e \leq 2^{30} -
+        1`; plus the special values ``+0``, ``-0``, ``+infinity``,
+        ``-infinity``, and ``NaN`` (which stands for Not-a-Number).
+
+        This function returns `s`, `m`, and `e-p`.  For the special values:
+
+        - ``+0`` returns ``(1, 0, 0)``
+        - ``-0`` returns ``(-1, 0, 0)``
+        - the return values for ``+infinity``, ``-infinity``, and ``NaN`` are
+          not specified.
+
+        EXAMPLES::
+
+            sage: # needs sage.symbolic
+            sage: a = RDF(exp(1.0)); a
+            2.718281828459045
+            sage: sign, mantissa, exponent = RDF(exp(1.0)).sign_mantissa_exponent()
+            sage: sign, mantissa, exponent
+            (1, 6121026514868073, -51)
+            sage: sign*mantissa*(2**exponent) == a
+            True
+
+        The mantissa is always a nonnegative number::
+
+            sage: RDF(-1).sign_mantissa_exponent()                                      # needs sage.rings.real_mpfr
+            (-1, 4503599627370496, -52)
+
+        TESTS::
+
+            sage: RDF('+0').sign_mantissa_exponent()                                    # needs sage.rings.real_mpfr
+            (1, 0, 0)
+            sage: RDF('-0').sign_mantissa_exponent()                                    # needs sage.rings.real_mpfr
+            (-1, 0, 0)
+        """
+        from sage.rings.real_mpfr import RR
+        return RR(self._value).sign_mantissa_exponent()
+
+    def as_integer_ratio(self):
+        """
+        Return a coprime pair of integers ``(a, b)`` such that ``self``
+        equals ``a / b`` exactly.
+
+        EXAMPLES::
+
+            sage: RDF(0).as_integer_ratio()
+            (0, 1)
+            sage: RDF(1/3).as_integer_ratio()
+            (6004799503160661, 18014398509481984)
+            sage: RDF(37/16).as_integer_ratio()
+            (37, 16)
+            sage: RDF(3^60).as_integer_ratio()
+            (42391158275216203520420085760, 1)
+        """
+        nd = float.as_integer_ratio(self._value)
+        return (Integer(nd[0]), Integer(nd[1]))
+
+    ########################
+    #   Basic Arithmetic
+    ########################
+    def __invert__(self):
+        """
+        Compute the multiplicative inverse of ``self``.
+
+        EXAMPLES::
+
+            sage: a = RDF(-1.5)*RDF(2.5)
+            sage: a.__invert__()
+            -0.26666666666666666
+            sage: ~a
+            -0.26666666666666666
+        """
+        cdef RealDoubleElement x = <RealDoubleElement>PY_NEW(RealDoubleElement)
+        x._value = 1.0 / self._value
+        return x
+
+    cpdef _add_(self, right):
+        """
+        Add two real numbers with the same parent.
+
+        EXAMPLES::
+
+            sage: RDF('-1.5') + RDF('2.5')  # indirect doctest
+            1.0
+        """
+        cdef RealDoubleElement x = <RealDoubleElement>PY_NEW(RealDoubleElement)
+        x._value = self._value + (<RealDoubleElement>right)._value
+        return x
+
+    cpdef _sub_(self, right):
+        """
+        Subtract two real numbers with the same parent.
+
+        EXAMPLES::
+
+            sage: RDF('-1.5') - RDF('2.5')  # indirect doctest
+            -4.0
+        """
+        cdef RealDoubleElement x = <RealDoubleElement>PY_NEW(RealDoubleElement)
+        x._value = self._value - (<RealDoubleElement>right)._value
+        return x
+
+    cpdef _mul_(self, right):
+        """
+        Multiply two real numbers with the same parent.
+
+        EXAMPLES::
+
+            sage: RDF('-1.5') * RDF('2.5')  # indirect doctest
+            -3.75
+        """
+        cdef RealDoubleElement x = <RealDoubleElement>PY_NEW(RealDoubleElement)
+        x._value = self._value * (<RealDoubleElement>right)._value
+        return x
+
+    cpdef _div_(self, right):
+        """
+        Divide ``self`` by ``right``.
+
+        EXAMPLES::
+
+            sage: RDF('-1.5') / RDF('2.5')  # indirect doctest
+            -0.6
+            sage: RDF(1)/RDF(0)
+            +infinity
+        """
+        cdef RealDoubleElement x = <RealDoubleElement>PY_NEW(RealDoubleElement)
+        x._value = self._value / (<RealDoubleElement>right)._value
+        return x
+
+    def __neg__(self):
+        """
+        Negate ``self``.
+
+        EXAMPLES::
+
+            sage: -RDF('-1.5')
+            1.5
+        """
+        cdef RealDoubleElement x = <RealDoubleElement>PY_NEW(RealDoubleElement)
+        x._value = -self._value
+        return x
+
+    def conjugate(self):
+        r"""
+        Return the complex conjugate of this real number, which is
+        the real number itself.
+
+        EXAMPLES::
+
+            sage: RDF(4).conjugate()
+            4.0
+        """
+        return self
+
+    def __abs__(self):
+        """
+        Return the absolute value of ``self``.
+
+        EXAMPLES::
+
+            sage: abs(RDF(1.5))
+            1.5
+            sage: abs(RDF(-1.5))
+            1.5
+            sage: abs(RDF(0.0))
+            0.0
+            sage: abs(RDF(-0.0))
+            0.0
+        """
+        # Use signbit instead of >= to handle -0.0 correctly
+        if not libc.math.signbit(self._value):
+            return self
+        else:
+            return self._new_c(-self._value)
+
+    cpdef RealDoubleElement abs(RealDoubleElement self):
+        """
+        Return the absolute value of ``self``.
+
+        EXAMPLES::
+
+            sage: RDF(1e10).abs()
+            10000000000.0
+            sage: RDF(-1e10).abs()
+            10000000000.0
+        """
+        if self._value >= 0:
+            return self
+        else:
+            return self._new_c(-self._value)
+
+    def __lshift__(x, y):
+        """
+        LShifting a double is not supported; nor is lshifting a
+        :class:`RealDoubleElement`.
+
+        TESTS::
+
+            sage: RDF(2) << 3
+            Traceback (most recent call last):
+            ...
+            TypeError: unsupported operand type(s) for <<
+        """
+        raise TypeError("unsupported operand type(s) for <<")
+
+    def __rshift__(x, y):
+        """
+        RShifting a double is not supported; nor is rshifting a
+        :class:`RealDoubleElement`.
+
+        TESTS::
+
+            sage: RDF(2) >> 3
+            Traceback (most recent call last):
+            ...
+            TypeError: unsupported operand type(s) for >>
+        """
+        raise TypeError("unsupported operand type(s) for >>")
+
+    def multiplicative_order(self):
+        r"""
+        Return `n` such that ``self^n == 1``.
+
+        Only `\pm 1` have finite multiplicative order.
+
+        EXAMPLES::
+
+            sage: RDF(1).multiplicative_order()
+            1
+            sage: RDF(-1).multiplicative_order()
+            2
+            sage: RDF(3).multiplicative_order()
+            +Infinity
+        """
+        if self._value == 1:
+            return 1
+        elif self._value == -1:
+            return 2
+        return sage.rings.infinity.infinity
+
+    def sign(self):
+        """
+        Return -1, 0, or 1 if ``self`` is negative, zero, or positive;
+        respectively.
+
+        EXAMPLES::
+
+            sage: RDF(-1.5).sign()
+            -1
+            sage: RDF(0).sign()
+            0
+            sage: RDF(2.5).sign()
+            1
+        """
+        if not self._value:
+            return 0
+        if self._value > 0:
+            return 1
+        return -1
+
+    ###################
+    # Rounding etc
+    ###################
+
+    def round(self):
+        """
+        Round ``self`` to the nearest integer.
+
+        This uses the convention of rounding half to even
+        (i.e., if the fractional part of ``self`` is `0.5`, then it
+        is rounded to the nearest even integer).
+
+        EXAMPLES::
+
+            sage: RDF(0.49).round()
+            0
+            sage: a=RDF(0.51).round(); a
+            1
+            sage: RDF(0.5).round()
+            0
+            sage: RDF(1.5).round()
+            2
+        """
+        return Integer(round(self._value))
+
+    def floor(self):
+        """
+        Return the floor of ``self``.
+
+        EXAMPLES::
+
+            sage: RDF(2.99).floor()
+            2
+            sage: RDF(2.00).floor()
+            2
+            sage: RDF(-5/2).floor()
+            -3
+        """
+        return Integer(math.floor(self._value))
+
+    def ceil(self):
+        """
+        Return the ceiling of ``self``.
+
+        EXAMPLES::
+
+            sage: RDF(2.99).ceil()
+            3
+            sage: RDF(2.00).ceil()
+            2
+            sage: RDF(-5/2).ceil()
+            -2
+        """
+        return Integer(math.ceil(self._value))
+
+    ceiling = ceil
+
+    def trunc(self):
+        """
+        Truncates this number (returns integer part).
+
+        EXAMPLES::
+
+            sage: RDF(2.99).trunc()
+            2
+            sage: RDF(-2.00).trunc()
+            -2
+            sage: RDF(0.00).trunc()
+            0
+        """
+        return Integer(int(self._value))
+
+    def frac(self):
+        """
+        Return a real number in `(-1, 1)`. It satisfies the relation:
+        ``x = x.trunc() + x.frac()``
+
+        EXAMPLES::
+
+            sage: RDF(2.99).frac()
+            0.9900000000000002
+            sage: RDF(2.50).frac()
+            0.5
+            sage: RDF(-2.79).frac()
+            -0.79
+        """
+        return self._new_c(self._value - int(self._value))
+
+    ###########################################
+    # Conversions
+    ###########################################
+
+    def __float__(self):
+        """
+        Return ``self`` as a python float.
+
+        EXAMPLES::
+
+            sage: float(RDF(1.5))
+            1.5
+            sage: type(float(RDF(1.5)))
+            <... 'float'>
+        """
+        return self._value
+
+    def _rpy_(self):
+        """
+        Return ``self.__float__()`` for rpy to convert into the
+        appropriate R object.
+
+        EXAMPLES::
+
+            sage: n = RDF(2.0)
+            sage: n._rpy_()
+            2.0
+            sage: type(n._rpy_())
+            <... 'float'>
+        """
+        return self.__float__()
+
+    def __int__(self):
+        """
+        Return integer truncation of this real number.
+
+        EXAMPLES::
+
+            sage: int(RDF(2.99))
+            2
+            sage: int(RDF(-2.99))
+            -2
+        """
+        return int(self._value)
+
+    def _complex_mpfr_field_(self, CC):
+        """
+        EXAMPLES::
+
+            sage: a = RDF(1/3)
+            sage: CC(a)                                                                 # needs sage.rings.real_mpfr
+            0.333333333333333
+            sage: a._complex_mpfr_field_(CC)                                            # needs sage.rings.real_mpfr
+            0.333333333333333
+
+        If we coerce to a higher-precision field the extra bits appear
+        random; they are actually 0s in base 2.
+
+        ::
+
+            sage: a._complex_mpfr_field_(ComplexField(100))                             # needs sage.rings.real_mpfr
+            0.33333333333333331482961625625
+            sage: a._complex_mpfr_field_(ComplexField(100)).str(2)                      # needs sage.rings.real_mpfr
+            '0.01010101010101010101010101010101010101010101010101010100000000000000000000000000000000000000000000000'
+        """
+        return CC(self._value)
+
+    def _complex_double_(self, CDF):
+        """
+        Return ``self`` as a complex double.
+
+        EXAMPLES::
+
+            sage: CDF(RDF(1/3))  # indirect doctest                                     # needs sage.rings.complex_double
+            0.3333333333333333
+        """
+        return CDF(self._value)
+
+    def __pari__(self):
+        """
+        Return a PARI representation of ``self``.
+
+        EXAMPLES::
+
+            sage: RDF(1.5).__pari__()                                                   # needs sage.libs.pari
+            1.50000000000000
+        """
+        global new_gen_from_real_double_element
+        if new_gen_from_real_double_element is None:
+            from sage.libs.pari.convert_sage_real_double import new_gen_from_real_double_element
+        return new_gen_from_real_double_element(self)
+
+    ###########################################
+    # Comparisons: ==, !=, <, <=, >, >=
+    ###########################################
+
+    def is_NaN(self):
+        """
+        Check if ``self`` is ``NaN``.
+
+        EXAMPLES::
+
+            sage: RDF(1).is_NaN()
+            False
+            sage: a = RDF(0)/RDF(0)
+            sage: a.is_NaN()
+            True
+        """
+        return bool(libc.math.isnan(self._value))
+
+    def is_positive_infinity(self):
+        r"""
+        Check if ``self`` is `+\infty`.
+
+        EXAMPLES::
+
+            sage: a = RDF(1)/RDF(0)
+            sage: a.is_positive_infinity()
+            True
+            sage: a = RDF(-1)/RDF(0)
+            sage: a.is_positive_infinity()
+            False
+        """
+        if not libc.math.isinf(self._value):
+            return False
+        return self._value > 0
+
+    def is_negative_infinity(self):
+        r"""
+        Check if ``self`` is `-\infty`.
+
+        EXAMPLES::
+
+            sage: a = RDF(2)/RDF(0)
+            sage: a.is_negative_infinity()
+            False
+            sage: a = RDF(-3)/RDF(0)
+            sage: a.is_negative_infinity()
+            True
+        """
+        if not libc.math.isinf(self._value):
+            return False
+        return self._value < 0
+
+    def is_infinity(self):
+        r"""
+        Check if ``self`` is `\infty`.
+
+        EXAMPLES::
+
+            sage: a = RDF(2); b = RDF(0)
+            sage: (a/b).is_infinity()
+            True
+            sage: (b/a).is_infinity()
+            False
+        """
+        return bool(libc.math.isinf(self._value))
+
+    cpdef _richcmp_(left, right, int op):
+        """
+        Rich comparison of ``left`` and ``right``.
+
+        EXAMPLES::
+
+            sage: RDF(2) < RDF(0)
+            False
+            sage: RDF(2) == RDF(4/2)
+            True
+            sage: RDF(-2) > RDF(-4)
+            True
+
+        TESTS:
+
+        Check comparisons with ``NaN`` (:issue:`16515`)::
+
+            sage: n = RDF('NaN')
+            sage: n == n
+            False
+            sage: n == RDF(1)
+            False
+        """
+        # We really need to use the correct operators, to deal
+        # correctly with NaNs.
+        cdef double x = (<RealDoubleElement>left)._value
+        cdef double y = (<RealDoubleElement>right)._value
+        if op == Py_LT:
+            return x < y
+        elif op == Py_LE:
+            return x <= y
+        elif op == Py_EQ:
+            return x == y
+        elif op == Py_NE:
+            return x != y
+        elif op == Py_GT:
+            return x > y
+        else:
+            return x >= y
+
+    ############################
+    # Special Functions
+    ############################
+
+    def NaN(self):
+        """
+        Return Not-a-Number ``NaN``.
+
+        EXAMPLES::
+
+            sage: RDF.NaN()
+            NaN
+        """
+        return self(0)/self(0)
+
+    nan = NaN
+
+    def sqrt(self, extend=True, all=False):
+        """
+        The square root function.
+
+        INPUT:
+
+        - ``extend`` -- boolean (default: ``True``); if ``True``, return a
+          square root in a complex field if necessary if ``self`` is negative.
+          Otherwise raise a :exc:`ValueError`.
+
+        - ``all`` -- boolean (default: ``False``); if ``True``, return a
+          list of all square roots
+
+        EXAMPLES::
+
+            sage: r = RDF(4.0)
+            sage: r.sqrt()
+            2.0
+            sage: r.sqrt()^2 == r
+            True
+
+        ::
+
+            sage: r = RDF(4344)
+            sage: r.sqrt()
+            65.90902821313632
+            sage: r.sqrt()^2 - r
+            0.0
+
+        ::
+
+            sage: r = RDF(-2.0)
+            sage: r.sqrt()                                                              # needs sage.rings.complex_double
+            1.4142135623730951*I
+
+        ::
+
+            sage: RDF(2).sqrt(all=True)
+            [1.4142135623730951, -1.4142135623730951]
+            sage: RDF(0).sqrt(all=True)
+            [0.0]
+            sage: RDF(-2).sqrt(all=True)                                                # needs sage.rings.complex_double
+            [1.4142135623730951*I, -1.4142135623730951*I]
+        """
+        if self._value >= 0:
+            x = self._new_c(libc.math.sqrt(self._value))
+            if all:
+                if x.is_zero():
+                    return [x]
+                else:
+                    return [x, -x]
+            else:
+                return x
+        if not extend:
+            raise ValueError("negative number %s does not have a square root in the real field" % self)
+        import sage.rings.complex_double
+        return self._complex_double_(sage.rings.complex_double.CDF).sqrt(all=all)
+
+    def is_square(self):
+        """
+        Return whether or not this number is a square in this field. For
+        the real numbers, this is ``True`` if and only if ``self`` is
+        nonnegative.
+
+        EXAMPLES::
+
+            sage: RDF(3.5).is_square()
+            True
+            sage: RDF(0).is_square()
+            True
+            sage: RDF(-4).is_square()
+            False
+        """
+        return self._value >= 0
+
+    def is_integer(self):
+        """
+        Return ``True`` if this number is a integer.
+
+        EXAMPLES::
+
+            sage: RDF(3.5).is_integer()
+            False
+            sage: RDF(3).is_integer()
+            True
+        """
+        return self._value in ZZ
+
+    def cube_root(self):
+        """
+        Return the cubic root (defined over the real numbers) of ``self``.
+
+        EXAMPLES::
+
+            sage: r = RDF(125.0); r.cube_root()
+            5.000000000000001
+            sage: r = RDF(-119.0)
+            sage: r.cube_root()^3 - r  # rel tol 1
+            -1.4210854715202004e-14
+        """
+        return self.nth_root(3)
+
+    def agm(self, other):
+        r"""
+        Return the arithmetic-geometric mean of ``self`` and ``other``. The
+        arithmetic-geometric mean is the common limit of the sequences
+        `u_n` and `v_n`, where `u_0` is ``self``,
+        `v_0` is other, `u_{n+1}` is the arithmetic mean
+        of `u_n` and `v_n`, and `v_{n+1}` is the
+        geometric mean of `u_n` and `v_n`. If any operand is negative, the
+        return value is ``NaN``.
+
+        EXAMPLES::
+
+            sage: a = RDF(1.5)
+            sage: b = RDF(2.3)
+            sage: a.agm(b)
+            1.8786484558146697
+
+        The arithmetic-geometric mean always lies between the geometric and
+        arithmetic mean::
+
+            sage: sqrt(a*b) < a.agm(b) < (a+b)/2
+            True
+        """
+        cdef double a = self._value
+        cdef double b = other
+        cdef double eps = 2.0**-51
+        if a < 0 or b < 0:
+            return self._parent.nan()
+        while True:
+            a1 = (a+b)/2
+            b1 = libc.math.sqrt(a*b)
+            if abs((b1/a1)-1) < eps:
+                return self._new_c(a1)
+            a, b = a1, b1
+
+    def algebraic_dependency(self, n):
+        """
+        Return a polynomial of degree at most `n` which is
+        approximately satisfied by this number.
+
+        .. NOTE::
+
+            The resulting polynomial need not be irreducible, and indeed
+            usually won't be if this number is a good approximation to an
+            algebraic number of degree less than `n`.
+
+        ALGORITHM:
+
+        Uses the PARI C-library :pari:`algdep` command.
+
+        EXAMPLES::
+
+            sage: r = sqrt(RDF(2)); r
+            1.4142135623730951
+            sage: r.algebraic_dependency(5)                                             # needs sage.libs.pari
+            x^2 - 2
+        """
+        return sage.arith.misc.algebraic_dependency(self, n)
+
+    algdep = algebraic_dependency
+
+cdef class ToRDF(Morphism):
+    def __init__(self, R):
+        """
+        Fast morphism from anything with a ``__float__`` method to an ``RDF``
+        element.
+
+        EXAMPLES::
+
+            sage: f = RDF.coerce_map_from(ZZ); f
+            Native morphism:
+              From: Integer Ring
+              To:   Real Double Field
+            sage: f(4)
+            4.0
+            sage: f = RDF.coerce_map_from(QQ); f
+            Native morphism:
+              From: Rational Field
+              To:   Real Double Field
+            sage: f(1/2)
+            0.5
+            sage: f = RDF.coerce_map_from(int); f
+            Native morphism:
+              From: Set of Python objects of class 'int'
+              To:   Real Double Field
+            sage: f(3r)
+            3.0
+            sage: f = RDF.coerce_map_from(float); f
+            Native morphism:
+              From: Set of Python objects of class 'float'
+              To:   Real Double Field
+            sage: f(3.5)
+            3.5
+        """
+        from sage.categories.homset import Hom
+        if isinstance(R, type):
+            from sage.sets.pythonclass import Set_PythonType
+            R = Set_PythonType(R)
+        Morphism.__init__(self, Hom(R, RDF))
+
+    cpdef Element _call_(self, x):
+        """
+        Send ``x`` to the image under this map.
+
+        EXAMPLES::
+
+            sage: f = RDF.coerce_map_from(float)
+            sage: f(3.5)  # indirect doctest
+            3.5
+        """
+        cdef RealDoubleElement r = <RealDoubleElement>PY_NEW(RealDoubleElement)
+        r._value = PyFloat_AsDouble(x)
+        return r
+
+    def _repr_type(self):
+        """
+        Return the representation type of ``self``.
+
+        EXAMPLES::
+
+            sage: RDF.coerce_map_from(float)._repr_type()
+            'Native'
+        """
+        return "Native"
+
+
+#####################################################
+# unique objects
+#####################################################
+cdef RealDoubleField_class _RDF
+_RDF = RealDoubleField_class()
+
+RDF = _RDF   # external interface
+
+
+def RealDoubleField():
+    """
+    Return the unique instance of the
+    :class:`real double field<RealDoubleField_class>`.
+
+    EXAMPLES::
+
+        sage: RealDoubleField() is RealDoubleField()
+        True
+    """
+    global _RDF
+    return _RDF
+
+
+def is_RealDoubleElement(x):
+    """
+    Check if ``x`` is an element of the real double field.
+
+    EXAMPLES::
+
+        sage: from sage.rings.real_double import is_RealDoubleElement
+        sage: is_RealDoubleElement(RDF(3))
+        doctest:warning...
+        DeprecationWarning: The function is_RealDoubleElement is deprecated;
+        use 'isinstance(..., RealDoubleElement)' instead.
+        See https://github.com/sagemath/sage/issues/38128 for details.
+        True
+        sage: is_RealDoubleElement(RIF(3))                                              # needs sage.rings.real_interval_field
+        False
+    """
+    from sage.misc.superseded import deprecation_cython
+    deprecation_cython(38128,
+                       "The function is_RealDoubleElement is deprecated; "
+                       "use 'isinstance(..., RealDoubleElement)' instead.")
+    return isinstance(x, RealDoubleElement)
+
+
+# ################ FAST CREATION CODE ######################
+#            Based on fast integer creation code
+#         There is nothing to see here, move along
+
+# We use a global element to steal all the references
+# from.  DO NOT INITIALIZE IT AGAIN and DO NOT REFERENCE IT!
+cdef RealDoubleElement global_dummy_element
+
+try:
+    from sage.rings.real_double_element_gsl import RealDoubleElement_gsl
+except ImportError:
+    global_dummy_element = RealDoubleElement(0)
+else:
+    global_dummy_element = RealDoubleElement_gsl(0)
+
+# A global pool for performance when elements are rapidly created and destroyed.
+# It operates on the following principles:
+#
+# - The pool starts out empty.
+# - When a new element is needed, one from the pool is returned
+#   if available, otherwise a new RealDoubleElement object is created
+# - When an element is collected, it will add it to the pool
+#   if there is room, otherwise it will be deallocated.
+DEF element_pool_size = 50
+
+cdef PyObject* element_pool[element_pool_size]
+cdef int element_pool_count = 0
+
+# used for profiling the pool
+cdef int total_alloc = 0
+cdef int use_pool = 0
+
+
+cdef PyObject* fast_tp_new(type t, args, kwds) noexcept:
+    global element_pool, element_pool_count, total_alloc, use_pool
+
+    cdef PyObject* new
+
+    # for profiling pool usage
+    # total_alloc += 1
+
+    # If there is a ready real double in the pool, we will
+    # decrement the counter and return that.
+
+    if element_pool_count > 0:
+
+        # for profiling pool usage
+        # use_pool += 1
+
+        element_pool_count -= 1
+        new = <PyObject *> element_pool[element_pool_count]
+
+    # Otherwise, we have to create one.
+
+    else:
+
+        # allocate enough room for the RealDoubleElement,
+        # sizeof_RealDoubleElement is sizeof(RealDoubleElement).
+        # The use of PyObject_Malloc directly assumes
+        # that RealDoubleElements are not garbage collected, i.e.
+        # they do not possess references to other Python
+        # objects (As indicated by the Py_TPFLAGS_HAVE_GC flag).
+        # See below for a more detailed description.
+
+        new = <PyObject*>PyObject_Malloc(sizeof(RealDoubleElement))
+
+        # Now set every member as set in z, the global dummy RealDoubleElement
+        # created before this tp_new started to operate.
+
+        memcpy(new, (<void*>global_dummy_element), sizeof(RealDoubleElement))
+
+    # This line is only needed if Python is compiled in debugging mode
+    # './configure --with-pydebug' or SAGE_DEBUG=yes. If that is the
+    # case a Python object has a bunch of debugging fields which are
+    # initialized with this macro.
+    if_Py_TRACE_REFS_then_PyObject_INIT(
+            new, Py_TYPE(global_dummy_element))
+
+    # The global_dummy_element may have a reference count larger than
+    # one, but it is expected that newly created objects have a
+    # reference count of one. This is potentially unneeded if
+    # everybody plays nice, because the global_dummy_element has only
+    # one reference in that case.
+
+    # Objects from the pool have reference count zero, so this
+    # needs to be set in this case.
+
+    new.ob_refcnt = 1
+
+    return new
+
+cdef void fast_tp_dealloc(PyObject* o) noexcept:
+
+    # If there is room in the pool for a used integer object,
+    # then put it in rather than deallocating it.
+
+    global element_pool, element_pool_count
+
+    if element_pool_count < element_pool_size:
+
+        # And add it to the pool.
+        element_pool[element_pool_count] = o
+        element_pool_count += 1
+        return
+
+    # Free the object. This assumes that Py_TPFLAGS_HAVE_GC is not
+    # set. If it was set another free function would need to be
+    # called.
+
+    PyObject_Free(o)
+
+
+from sage.misc.allocator cimport hook_tp_functions, hook_tp_functions_type
+hook_tp_functions(global_dummy_element, <newfunc>(&fast_tp_new), <destructor>(&fast_tp_dealloc), False)
+try:
+    from sage.rings.real_double_element_gsl import RealDoubleElement_gsl
+except Exception:
+    pass
+else:
+    # global_dummy_element is of type RealDoubleElement_gsl,
+    # so hook the base class now.
+    hook_tp_functions_type(RealDoubleElement, <newfunc>(&fast_tp_new), <destructor>(&fast_tp_dealloc), False)
+    # From here on, calling PY_NEW(RealDoubleElement) actually creates an instance of RealDoubleElement_gsl
+
+
+cdef double_repr(double x):
+    """
+    Convert a double to a string with maximum precision.
+    """
+    if libc.math.isfinite(x):
+        return repr(x)
+    if libc.math.isinf(x):
+        if x > 0:
+            return "+infinity"
+        if x < 0:
+            return "-infinity"
+    return "NaN"
+
+
+# Support Python's numbers abstract base class
+import numbers
+numbers.Real.register(RealDoubleElement)