Question

I came across this while debugging some unit tests that compared a returned single array against the expected results.

System.Single.NaN == System.Single.Nan;

The unit test was expecting true, but this evaluates to false. Why does this evaluate to false when the other static methods of single return true? Is NaN not a constant value ?

System.Single.MaxValue == System.Single.MaxValue;
System.Single.Epsilon == System.Single.Epsilon;
null == null;

http://msdn.microsoft.com/en-us/library/system.single.nan.aspx states: “Two NaN values are considered unequal to one another.” but this question is out of curiosity as to why this is so more than anything else.

Answer 1

Because the IEEE 754 standard (the one used for float and double) says so.

To quote the wiki

A comparison with a NaN always returns an unordered result even when comparing with itself. The comparison predicates are either signaling or non-signaling, the signaling versions signal an invalid exception for such comparisons. The equality and inequality predicates are non-signaling so x = x returning false can be used to test if x is a quiet NaN. The other standard comparison predicates are all signaling if they receive a NaN operand, the standard also provides non-signaling versions of these other predicates. The predicate isNaN(x) determines if a value is a NaN and never signals an exception, even if x is a signaling NaN.

(note that .NET doesn’t support the signaling NaN, and probably treats it as non-signaling)