比较两个双精度浮点数或两个浮点数最有效的方法是什么?
简单地这样做是不正确的:
bool CompareDoubles1 (double A, double B)
{
return A == B;
}
比如:
bool CompareDoubles2 (double A, double B)
{
diff = A - B;
return (diff < EPSILON) && (-diff < EPSILON);
}
似乎是浪费加工。
有人知道更聪明的浮点比较器吗?
当前回答
你必须为浮点数比较做这个处理，因为浮点数不能像整数类型那样完美地比较。下面是各种比较运算符的函数。
浮点数等于(==)
我也更喜欢减法技术，而不是依赖于fabs()或abs()，但我必须在从64位PC到ATMega328微控制器(Arduino)的各种架构上快速配置它，才能真正看到它是否会产生很大的性能差异。
所以，让我们忘记这些绝对值的东西，只做一些减法和比较!
从微软的例子修改如下:
/// @brief See if two floating point numbers are approximately equal.
/// @param[in] a number 1
/// @param[in] b number 2
/// @param[in] epsilon A small value such that if the difference between the two numbers is
/// smaller than this they can safely be considered to be equal.
/// @return true if the two numbers are approximately equal, and false otherwise
bool is_float_eq(float a, float b, float epsilon) {
return ((a - b) < epsilon) && ((b - a) < epsilon);
}
bool is_double_eq(double a, double b, double epsilon) {
return ((a - b) < epsilon) && ((b - a) < epsilon);
}
使用示例:
constexpr float EPSILON = 0.0001; // 1e-4
is_float_eq(1.0001, 0.99998, EPSILON);
我不完全确定，但在我看来，对基于epsilon的方法的一些批评，正如这个高度好评的答案下面的评论所描述的那样，可以通过使用变量epsilon来解决，根据比较的浮点值缩放，像这样:
float a = 1.0001;
float b = 0.99998;
float epsilon = std::max(std::fabs(a), std::fabs(b)) * 1e-4;
is_float_eq(a, b, epsilon);
通过这种方式，epsilon值随浮点值伸缩，因此它的值不会小到不重要。
为了完整起见，让我们添加剩下的:
大于(>)小于(<):
/// @brief See if floating point number `a` is > `b`
/// @param[in] a number 1
/// @param[in] b number 2
/// @param[in] epsilon a small value such that if `a` is > `b` by this amount, `a` is considered
/// to be definitively > `b`
/// @return true if `a` is definitively > `b`, and false otherwise
bool is_float_gt(float a, float b, float epsilon) {
return a > b + epsilon;
}
bool is_double_gt(double a, double b, double epsilon) {
return a > b + epsilon;
}
/// @brief See if floating point number `a` is < `b`
/// @param[in] a number 1
/// @param[in] b number 2
/// @param[in] epsilon a small value such that if `a` is < `b` by this amount, `a` is considered
/// to be definitively < `b`
/// @return true if `a` is definitively < `b`, and false otherwise
bool is_float_lt(float a, float b, float epsilon) {
return a < b - epsilon;
}
bool is_double_lt(double a, double b, double epsilon) {
return a < b - epsilon;
}
大于或等于(>=)，小于或等于(<=)
/// @brief Returns true if `a` is definitively >= `b`, and false otherwise
bool is_float_ge(float a, float b, float epsilon) {
return a > b - epsilon;
}
bool is_double_ge(double a, double b, double epsilon) {
return a > b - epsilon;
}
/// @brief Returns true if `a` is definitively <= `b`, and false otherwise
bool is_float_le(float a, float b, float epsilon) {
return a < b + epsilon;
}
bool is_double_le(double a, double b, double epsilon) {
return a < b + epsilon;
}
额外的改进:
A good default value for epsilon in C++ is std::numeric_limits<T>::epsilon(), which evaluates to either 0 or FLT_EPSILON, DBL_EPSILON, or LDBL_EPSILON. See here: https://en.cppreference.com/w/cpp/types/numeric_limits/epsilon. You can also see the float.h header for FLT_EPSILON, DBL_EPSILON, and LDBL_EPSILON. See https://en.cppreference.com/w/cpp/header/cfloat and https://www.cplusplus.com/reference/cfloat/ You could template the functions instead, to handle all floating point types: float, double, and long double, with type checks for these types via a static_assert() inside the template. Scaling the epsilon value is a good idea to ensure it works for really large and really small a and b values. This article recommends and explains it: http://realtimecollisiondetection.net/blog/?p=89. So, you should scale epsilon by a scaling value equal to max(1.0, abs(a), abs(b)), as that article explains. Otherwise, as a and/or b increase in magnitude, the epsilon would eventually become so small relative to those values that it becomes lost in the floating point error. So, we scale it to become larger in magnitude like they are. However, using 1.0 as the smallest allowed scaling factor for epsilon also ensures that for really small-magnitude a and b values, epsilon itself doesn't get scaled so small that it also becomes lost in the floating point error. So, we limit the minimum scaling factor to 1.0. If you want to "encapsulate" the above functions into a class, don't. Instead, wrap them up in a namespace if you like in order to namespace them. Ex: if you put all of the stand-alone functions into a namespace called float_comparison, then you could access the is_eq() function like this, for instance: float_comparison::is_eq(1.0, 1.5);. It might also be nice to add comparisons against zero, not just comparisons between two values. So, here is a better type of solution with the above improvements in place: namespace float_comparison { /// Scale the epsilon value to become large for large-magnitude a or b, /// but no smaller than 1.0, per the explanation above, to ensure that /// epsilon doesn't ever fall out in floating point error as a and/or b /// increase in magnitude. template<typename T> static constexpr T scale_epsilon(T a, T b, T epsilon = std::numeric_limits<T>::epsilon()) noexcept { static_assert(std::is_floating_point_v<T>, "Floating point comparisons " "require type float, double, or long double."); T scaling_factor; // Special case for when a or b is infinity if (std::isinf(a) || std::isinf(b)) { scaling_factor = 0; } else { scaling_factor = std::max({(T)1.0, std::abs(a), std::abs(b)}); } T epsilon_scaled = scaling_factor * std::abs(epsilon); return epsilon_scaled; } // Compare two values /// Equal: returns true if a is approximately == b, and false otherwise template<typename T> static constexpr bool is_eq(T a, T b, T epsilon = std::numeric_limits<T>::epsilon()) noexcept { static_assert(std::is_floating_point_v<T>, "Floating point comparisons " "require type float, double, or long double."); // test `a == b` first to see if both a and b are either infinity // or -infinity return a == b || std::abs(a - b) <= scale_epsilon(a, b, epsilon); } /* etc. etc.: is_eq() is_ne() is_lt() is_le() is_gt() is_ge() */ // Compare against zero /// Equal: returns true if a is approximately == 0, and false otherwise template<typename T> static constexpr bool is_eq_zero(T a, T epsilon = std::numeric_limits<T>::epsilon()) noexcept { static_assert(std::is_floating_point_v<T>, "Floating point comparisons " "require type float, double, or long double."); return is_eq(a, (T)0.0, epsilon); } /* etc. etc.: is_eq_zero() is_ne_zero() is_lt_zero() is_le_zero() is_gt_zero() is_ge_zero() */ } // namespace float_comparison
参见:
The macro forms of some of the functions above in my repo here: utilities.h. UPDATE 29 NOV 2020: it's a work-in-progress, and I'm going to make it a separate answer when ready, but I've produced a better, scaled-epsilon version of all of the functions in C in this file here: utilities.c. Take a look. ADDITIONAL READING I need to do now have done: Floating-point tolerances revisited, by Christer Ericson. VERY USEFUL ARTICLE! It talks about scaling epsilon in order to ensure it never falls out in floating point error, even for really large-magnitude a and/or b values!
其他回答
I found that the Google C++ Testing Framework contains a nice cross-platform template-based implementation of AlmostEqual2sComplement which works on both doubles and floats. Given that it is released under the BSD license, using it in your own code should be no problem, as long as you retain the license. I extracted the below code from http://code.google.com/p/googletest/source/browse/trunk/include/gtest/internal/gtest-internal.h https://github.com/google/googletest/blob/master/googletest/include/gtest/internal/gtest-internal.h and added the license on top.
一定要将GTEST_OS_WINDOWS定义为某个值(或者将使用它的代码更改为适合您的代码库的代码-毕竟它是BSD许可的)。
使用的例子:
double left = // something
double right = // something
const FloatingPoint<double> lhs(left), rhs(right);
if (lhs.AlmostEquals(rhs)) {
//they're equal!
}
代码如下:
// Copyright 2005, Google Inc.
// All rights reserved.
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
//
// Authors: wan@google.com (Zhanyong Wan), eefacm@gmail.com (Sean Mcafee)
//
// The Google C++ Testing Framework (Google Test)
// This template class serves as a compile-time function from size to
// type. It maps a size in bytes to a primitive type with that
// size. e.g.
//
// TypeWithSize<4>::UInt
//
// is typedef-ed to be unsigned int (unsigned integer made up of 4
// bytes).
//
// Such functionality should belong to STL, but I cannot find it
// there.
//
// Google Test uses this class in the implementation of floating-point
// comparison.
//
// For now it only handles UInt (unsigned int) as that's all Google Test
// needs. Other types can be easily added in the future if need
// arises.
template <size_t size>
class TypeWithSize {
public:
// This prevents the user from using TypeWithSize<N> with incorrect
// values of N.
typedef void UInt;
};
// The specialization for size 4.
template <>
class TypeWithSize<4> {
public:
// unsigned int has size 4 in both gcc and MSVC.
//
// As base/basictypes.h doesn't compile on Windows, we cannot use
// uint32, uint64, and etc here.
typedef int Int;
typedef unsigned int UInt;
};
// The specialization for size 8.
template <>
class TypeWithSize<8> {
public:
#if GTEST_OS_WINDOWS
typedef __int64 Int;
typedef unsigned __int64 UInt;
#else
typedef long long Int; // NOLINT
typedef unsigned long long UInt; // NOLINT
#endif // GTEST_OS_WINDOWS
};
// This template class represents an IEEE floating-point number
// (either single-precision or double-precision, depending on the
// template parameters).
//
// The purpose of this class is to do more sophisticated number
// comparison. (Due to round-off error, etc, it's very unlikely that
// two floating-points will be equal exactly. Hence a naive
// comparison by the == operation often doesn't work.)
//
// Format of IEEE floating-point:
//
// The most-significant bit being the leftmost, an IEEE
// floating-point looks like
//
// sign_bit exponent_bits fraction_bits
//
// Here, sign_bit is a single bit that designates the sign of the
// number.
//
// For float, there are 8 exponent bits and 23 fraction bits.
//
// For double, there are 11 exponent bits and 52 fraction bits.
//
// More details can be found at
// http://en.wikipedia.org/wiki/IEEE_floating-point_standard.
//
// Template parameter:
//
// RawType: the raw floating-point type (either float or double)
template <typename RawType>
class FloatingPoint {
public:
// Defines the unsigned integer type that has the same size as the
// floating point number.
typedef typename TypeWithSize<sizeof(RawType)>::UInt Bits;
// Constants.
// # of bits in a number.
static const size_t kBitCount = 8*sizeof(RawType);
// # of fraction bits in a number.
static const size_t kFractionBitCount =
std::numeric_limits<RawType>::digits - 1;
// # of exponent bits in a number.
static const size_t kExponentBitCount = kBitCount - 1 - kFractionBitCount;
// The mask for the sign bit.
static const Bits kSignBitMask = static_cast<Bits>(1) << (kBitCount - 1);
// The mask for the fraction bits.
static const Bits kFractionBitMask =
~static_cast<Bits>(0) >> (kExponentBitCount + 1);
// The mask for the exponent bits.
static const Bits kExponentBitMask = ~(kSignBitMask | kFractionBitMask);
// How many ULP's (Units in the Last Place) we want to tolerate when
// comparing two numbers. The larger the value, the more error we
// allow. A 0 value means that two numbers must be exactly the same
// to be considered equal.
//
// The maximum error of a single floating-point operation is 0.5
// units in the last place. On Intel CPU's, all floating-point
// calculations are done with 80-bit precision, while double has 64
// bits. Therefore, 4 should be enough for ordinary use.
//
// See the following article for more details on ULP:
// http://www.cygnus-software.com/papers/comparingfloats/comparingfloats.htm.
static const size_t kMaxUlps = 4;
// Constructs a FloatingPoint from a raw floating-point number.
//
// On an Intel CPU, passing a non-normalized NAN (Not a Number)
// around may change its bits, although the new value is guaranteed
// to be also a NAN. Therefore, don't expect this constructor to
// preserve the bits in x when x is a NAN.
explicit FloatingPoint(const RawType& x) { u_.value_ = x; }
// Static methods
// Reinterprets a bit pattern as a floating-point number.
//
// This function is needed to test the AlmostEquals() method.
static RawType ReinterpretBits(const Bits bits) {
FloatingPoint fp(0);
fp.u_.bits_ = bits;
return fp.u_.value_;
}
// Returns the floating-point number that represent positive infinity.
static RawType Infinity() {
return ReinterpretBits(kExponentBitMask);
}
// Non-static methods
// Returns the bits that represents this number.
const Bits &bits() const { return u_.bits_; }
// Returns the exponent bits of this number.
Bits exponent_bits() const { return kExponentBitMask & u_.bits_; }
// Returns the fraction bits of this number.
Bits fraction_bits() const { return kFractionBitMask & u_.bits_; }
// Returns the sign bit of this number.
Bits sign_bit() const { return kSignBitMask & u_.bits_; }
// Returns true iff this is NAN (not a number).
bool is_nan() const {
// It's a NAN if the exponent bits are all ones and the fraction
// bits are not entirely zeros.
return (exponent_bits() == kExponentBitMask) && (fraction_bits() != 0);
}
// Returns true iff this number is at most kMaxUlps ULP's away from
// rhs. In particular, this function:
//
// - returns false if either number is (or both are) NAN.
// - treats really large numbers as almost equal to infinity.
// - thinks +0.0 and -0.0 are 0 DLP's apart.
bool AlmostEquals(const FloatingPoint& rhs) const {
// The IEEE standard says that any comparison operation involving
// a NAN must return false.
if (is_nan() || rhs.is_nan()) return false;
return DistanceBetweenSignAndMagnitudeNumbers(u_.bits_, rhs.u_.bits_)
<= kMaxUlps;
}
private:
// The data type used to store the actual floating-point number.
union FloatingPointUnion {
RawType value_; // The raw floating-point number.
Bits bits_; // The bits that represent the number.
};
// Converts an integer from the sign-and-magnitude representation to
// the biased representation. More precisely, let N be 2 to the
// power of (kBitCount - 1), an integer x is represented by the
// unsigned number x + N.
//
// For instance,
//
// -N + 1 (the most negative number representable using
// sign-and-magnitude) is represented by 1;
// 0 is represented by N; and
// N - 1 (the biggest number representable using
// sign-and-magnitude) is represented by 2N - 1.
//
// Read http://en.wikipedia.org/wiki/Signed_number_representations
// for more details on signed number representations.
static Bits SignAndMagnitudeToBiased(const Bits &sam) {
if (kSignBitMask & sam) {
// sam represents a negative number.
return ~sam + 1;
} else {
// sam represents a positive number.
return kSignBitMask | sam;
}
}
// Given two numbers in the sign-and-magnitude representation,
// returns the distance between them as an unsigned number.
static Bits DistanceBetweenSignAndMagnitudeNumbers(const Bits &sam1,
const Bits &sam2) {
const Bits biased1 = SignAndMagnitudeToBiased(sam1);
const Bits biased2 = SignAndMagnitudeToBiased(sam2);
return (biased1 >= biased2) ? (biased1 - biased2) : (biased2 - biased1);
}
FloatingPointUnion u_;
};
编辑:这篇文章是4年前写的。它可能仍然有效，代码也很好，但有些人发现了改进。最好从谷歌Test源代码中获得最新版本的AlmostEquals，而不是我粘贴在这里的那个。
你必须为浮点数比较做这个处理，因为浮点数不能像整数类型那样完美地比较。下面是各种比较运算符的函数。
浮点数等于(==)
我也更喜欢减法技术，而不是依赖于fabs()或abs()，但我必须在从64位PC到ATMega328微控制器(Arduino)的各种架构上快速配置它，才能真正看到它是否会产生很大的性能差异。
所以，让我们忘记这些绝对值的东西，只做一些减法和比较!
从微软的例子修改如下:
/// @brief See if two floating point numbers are approximately equal.
/// @param[in] a number 1
/// @param[in] b number 2
/// @param[in] epsilon A small value such that if the difference between the two numbers is
/// smaller than this they can safely be considered to be equal.
/// @return true if the two numbers are approximately equal, and false otherwise
bool is_float_eq(float a, float b, float epsilon) {
return ((a - b) < epsilon) && ((b - a) < epsilon);
}
bool is_double_eq(double a, double b, double epsilon) {
return ((a - b) < epsilon) && ((b - a) < epsilon);
}
使用示例:
constexpr float EPSILON = 0.0001; // 1e-4
is_float_eq(1.0001, 0.99998, EPSILON);
我不完全确定，但在我看来，对基于epsilon的方法的一些批评，正如这个高度好评的答案下面的评论所描述的那样，可以通过使用变量epsilon来解决，根据比较的浮点值缩放，像这样:
float a = 1.0001;
float b = 0.99998;
float epsilon = std::max(std::fabs(a), std::fabs(b)) * 1e-4;
is_float_eq(a, b, epsilon);
通过这种方式，epsilon值随浮点值伸缩，因此它的值不会小到不重要。
为了完整起见，让我们添加剩下的:
大于(>)小于(<):
/// @brief See if floating point number `a` is > `b`
/// @param[in] a number 1
/// @param[in] b number 2
/// @param[in] epsilon a small value such that if `a` is > `b` by this amount, `a` is considered
/// to be definitively > `b`
/// @return true if `a` is definitively > `b`, and false otherwise
bool is_float_gt(float a, float b, float epsilon) {
return a > b + epsilon;
}
bool is_double_gt(double a, double b, double epsilon) {
return a > b + epsilon;
}
/// @brief See if floating point number `a` is < `b`
/// @param[in] a number 1
/// @param[in] b number 2
/// @param[in] epsilon a small value such that if `a` is < `b` by this amount, `a` is considered
/// to be definitively < `b`
/// @return true if `a` is definitively < `b`, and false otherwise
bool is_float_lt(float a, float b, float epsilon) {
return a < b - epsilon;
}
bool is_double_lt(double a, double b, double epsilon) {
return a < b - epsilon;
}
大于或等于(>=)，小于或等于(<=)
/// @brief Returns true if `a` is definitively >= `b`, and false otherwise
bool is_float_ge(float a, float b, float epsilon) {
return a > b - epsilon;
}
bool is_double_ge(double a, double b, double epsilon) {
return a > b - epsilon;
}
/// @brief Returns true if `a` is definitively <= `b`, and false otherwise
bool is_float_le(float a, float b, float epsilon) {
return a < b + epsilon;
}
bool is_double_le(double a, double b, double epsilon) {
return a < b + epsilon;
}
额外的改进:
A good default value for epsilon in C++ is std::numeric_limits<T>::epsilon(), which evaluates to either 0 or FLT_EPSILON, DBL_EPSILON, or LDBL_EPSILON. See here: https://en.cppreference.com/w/cpp/types/numeric_limits/epsilon. You can also see the float.h header for FLT_EPSILON, DBL_EPSILON, and LDBL_EPSILON. See https://en.cppreference.com/w/cpp/header/cfloat and https://www.cplusplus.com/reference/cfloat/ You could template the functions instead, to handle all floating point types: float, double, and long double, with type checks for these types via a static_assert() inside the template. Scaling the epsilon value is a good idea to ensure it works for really large and really small a and b values. This article recommends and explains it: http://realtimecollisiondetection.net/blog/?p=89. So, you should scale epsilon by a scaling value equal to max(1.0, abs(a), abs(b)), as that article explains. Otherwise, as a and/or b increase in magnitude, the epsilon would eventually become so small relative to those values that it becomes lost in the floating point error. So, we scale it to become larger in magnitude like they are. However, using 1.0 as the smallest allowed scaling factor for epsilon also ensures that for really small-magnitude a and b values, epsilon itself doesn't get scaled so small that it also becomes lost in the floating point error. So, we limit the minimum scaling factor to 1.0. If you want to "encapsulate" the above functions into a class, don't. Instead, wrap them up in a namespace if you like in order to namespace them. Ex: if you put all of the stand-alone functions into a namespace called float_comparison, then you could access the is_eq() function like this, for instance: float_comparison::is_eq(1.0, 1.5);. It might also be nice to add comparisons against zero, not just comparisons between two values. So, here is a better type of solution with the above improvements in place: namespace float_comparison { /// Scale the epsilon value to become large for large-magnitude a or b, /// but no smaller than 1.0, per the explanation above, to ensure that /// epsilon doesn't ever fall out in floating point error as a and/or b /// increase in magnitude. template<typename T> static constexpr T scale_epsilon(T a, T b, T epsilon = std::numeric_limits<T>::epsilon()) noexcept { static_assert(std::is_floating_point_v<T>, "Floating point comparisons " "require type float, double, or long double."); T scaling_factor; // Special case for when a or b is infinity if (std::isinf(a) || std::isinf(b)) { scaling_factor = 0; } else { scaling_factor = std::max({(T)1.0, std::abs(a), std::abs(b)}); } T epsilon_scaled = scaling_factor * std::abs(epsilon); return epsilon_scaled; } // Compare two values /// Equal: returns true if a is approximately == b, and false otherwise template<typename T> static constexpr bool is_eq(T a, T b, T epsilon = std::numeric_limits<T>::epsilon()) noexcept { static_assert(std::is_floating_point_v<T>, "Floating point comparisons " "require type float, double, or long double."); // test `a == b` first to see if both a and b are either infinity // or -infinity return a == b || std::abs(a - b) <= scale_epsilon(a, b, epsilon); } /* etc. etc.: is_eq() is_ne() is_lt() is_le() is_gt() is_ge() */ // Compare against zero /// Equal: returns true if a is approximately == 0, and false otherwise template<typename T> static constexpr bool is_eq_zero(T a, T epsilon = std::numeric_limits<T>::epsilon()) noexcept { static_assert(std::is_floating_point_v<T>, "Floating point comparisons " "require type float, double, or long double."); return is_eq(a, (T)0.0, epsilon); } /* etc. etc.: is_eq_zero() is_ne_zero() is_lt_zero() is_le_zero() is_gt_zero() is_ge_zero() */ } // namespace float_comparison
参见:
The macro forms of some of the functions above in my repo here: utilities.h. UPDATE 29 NOV 2020: it's a work-in-progress, and I'm going to make it a separate answer when ready, but I've produced a better, scaled-epsilon version of all of the functions in C in this file here: utilities.c. Take a look. ADDITIONAL READING I need to do now have done: Floating-point tolerances revisited, by Christer Ericson. VERY USEFUL ARTICLE! It talks about scaling epsilon in order to ensure it never falls out in floating point error, even for really large-magnitude a and/or b values!
不幸的是，即使您的“浪费”代码也是不正确的。EPSILON是可以添加到1.0并更改其值的最小值。值1.0非常重要——更大的数字在添加到EPSILON时不会改变。现在，您可以将这个值缩放到您正在比较的数字，以判断它们是否不同。比较两个双精度对象的正确表达式是:
if (fabs(a - b) <= DBL_EPSILON * fmax(fabs(a), fabs(b)))
{
// ...
}
这是最小值。一般来说，你会想要在计算中考虑噪声，并忽略一些最不重要的位，所以更现实的比较应该是这样的:
if (fabs(a - b) <= 16 * DBL_EPSILON * fmax(fabs(a), fabs(b)))
{
// ...
}
如果比较性能对您非常重要，并且您知道值的范围，那么您应该使用定点数字。
在这个版本中，你可以检查，这些数字之间的差异并不比某些分数(比如，0.0001%)更大:
bool floatApproximatelyEquals(const float a, const float b) {
if (b == 0.) return a == 0.; // preventing division by zero
return abs(1. - a / b) < 1e-6;
}
请注意Sneftel关于浮动可能的分数限制的评论。
还要注意的是，它不同于使用绝对的epsilon的方法——这里你不需要担心“数量级”——数字可能是，比如说1e100，或者1e-100，它们总是会被一致地比较，而且你不必为每一种情况更新epsilon。
你不能用一个固定的。根据double的值，EPSILON会发生变化。
更好的双比较应该是:
bool same(double a, double b)
{
return std::nextafter(a, std::numeric_limits<double>::lowest()) <= b
&& std::nextafter(a, std::numeric_limits<double>::max()) >= b;
}
