[Octave-bug-tracker] [bug #48897] Same value but not equal? (1.235*10^5 == 123500 -> ans = 0)

[anonymous]

URL:
  <http://savannah.gnu.org/bugs/?48897>

                 Summary: Same value but not equal? (1.235*10^5 == 123500 ->
ans = 0)
                 Project: GNU Octave
            Submitted by: None
            Submitted on: Thu Aug 25 10:26:34 2016
                Category: None
                Severity: 3 - Normal
                Priority: 5 - Normal
              Item Group: None
                  Status: None
             Assigned to: None
         Originator Name: 
        Originator Email: 
             Open/Closed: Open
         Discussion Lock: Any
                 Release: 4.0.3
        Operating System: Mac OS

    _______________________________________________________

Details:

Same value but not equal?

I found that octave works wrong with following specific value as follows:

>> x = 1.235*10^5
x =    1.2350e+05

>> y = 123500
y =  123500

>> x == y
ans = 0



Both x and y are double.
I found the difference between x and y after several trials:

>> typecast(x,"int8")
ans =

    1    0    0    0  -64   38   -2   64

>> typecast(y, "int8")
ans =

    0    0    0    0  -64   38   -2   64

The most significant bit of x is wrong here.
When I converted x to string and then reconverted to number, it turns OK.

>> str2num( num2str( x) ) == y
ans =  1

This is very weird. I hope this problem will be solved.
Thank you.




    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/bugs/?48897>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.gnu.org/