让我们考虑这三个数字：

A = 9007199254740992.
B = 9007199254740993.
C = 9007199254740992.0.

在实际规则下，A和C彼此相等，但不同于B：

> A == B.  %% A and B are both integers, compared as integers
false
> A == C.  %% C gets converted to an integer before comparison
true
> B == C.  %% C gets converted to an integer before comparison
false

如果是相反的方式，在与浮点数比较之前，将大于阈值的整数转换为浮点数，会怎么样？

> A == B.         %% no change, because they are both integers
false
> float(A) == C.  %% no surprise here
true
> float(B) == C.  %% B cannot be accurately represented as a floating point value!
true

所以现在看起来A和B都等于C，但彼此不相等，相等比较失去了传递性。
9007199254740992等于253，而53也是64位IEEE 754浮点数 * 中的有效位数，因此对于大于此值的数字，浮点类型不能表示所有整数。例如，9007199254740992.0 + 1 == 9007199254740992.0返回true。这就是Erlang整数类型（它是一个bignum，因此可以表示任意大的整数）被认为在此阈值之上更精确。

• 二进制表示法仅使用52位有效数，并且由于第一位几乎总是1，因此可以避免这种情况。搜索“次正规数”以了解更多信息。

Given that floats and integers have different precissions (floats have higher resolution with values closer to 0 and lower resolution with values further away from 0), if you want to compare them you need to transform one of them to the other.
If this transformation is from the one with higher precission to the one of lower precission, this would be possible:

H2 = H1 + Delta,
true = L == H1,
true = L == H2,
true = H1 /= H2. %% In this case, both H1 and H2 are equal to L but different between themselves

For this reason (transitivity), the transformation is performed the other way around, from lower precssion to higher precission.