numpy - python "is" returns True but "==" returns False -

"==" should not be returned unless true "gives" right? [101]: NPNAN NPNN is NPNAN Out [101]: Really [102]: NPNAN == NP. Nan out [102]: In the wrong [103]: NP. NaN == np.nan out [103]: False [104]: np.NaN == np.NAN out [104]: Incorrect

Edit:

What float Nain bus are the 3 expressions for the legacy of the old NMP versions or for their other uses?

No, for no you have found one of the exceptions. According to, the NAN is not equal to anything, not even by itself:

Compared with an NAN, always gives an anord result while comparing itself.

Not only behave like this:

  & gt; & Gt; & Gt; Nain = float ('nan')> gt; & Gt; & Gt; Nain hai nain true & gt; & Gt; & Gt; Nain == Nain false   
 As per the IEEE standard, there will be any programming language implementing floating point arithmetic.  
 Common Capitalization Name for Multiple Spelling. It is an abbreviation for  not a number  and individual people have briefly taken advantage of it briefly  
  is  always the same object Is going to be  true , but  ==  is also not guaranteed to return boolean. From:  
 
 Through the conference, for a successful comparison, the  incorrect  and  true  are returned, however, So if the comparison operator is used in the Boolean reference (for example,  if  is in the statement state), then python  bool ()  value To determine whether the result is true or false.