Modify ↓
Opened 10 years ago
Closed 10 years ago
#2810 closed Bug (No Bug)
handling of some float numbers
Reported by: | anonymous | Owned by: | |
---|---|---|---|
Milestone: | Component: | AutoIt | |
Version: | 3.3.12.0 | Severity: | None |
Keywords: | Cc: |
Description
wierd internal rounding.
$n1=0.9999999999999995 $n2=0.9999999999999996 $n3=0.99999999999999994449919099999 $n4=0.999999999999999944499191 ConsoleWrite('IsInt($n1)='&IsInt($n1)&'; $n1='&$n1&@cr) ;IsInt($n1)=0; $n1=0.999999999999999 ConsoleWrite('IsInt($n2)='&IsInt($n2)&'; $n2='&$n2&@cr) ;IsInt($n2)=0; $n2=1 ConsoleWrite('IsInt($n3)='&IsInt($n3)&'; $n3='&$n3&@cr) ;IsInt($n3)=0; $n3=1 ConsoleWrite('IsInt($n4)='&IsInt($n4)&'; $n4='&$n4&@cr) ;IsInt($n4)=1; $n4=1
Attachments (0)
Change History (1)
comment:1 Changed 10 years ago by jchd18
- Resolution set to No Bug
- Status changed from new to closed
Guidelines for posting comments:
- You cannot re-open a ticket but you may still leave a comment if you have additional information to add.
- In-depth discussions should take place on the forum.
For more information see the full version of the ticket guidelines here.
Note: See
TracTickets for help on using
tickets.
Again there is no bug here.
You can invent an infinite number of cases where the result of a floating point calculation differs from what you'd expect. Again it has nothing to do with AutoIt, which uses the relies on C++ compiler to provide "correct" results, "correct" being taken modulo what IEEE754 dictates. Please read http://en.wikipedia.org/wiki/Floating_point
IsInt($n) = 0 applied to a floating point value $n means that $n is not exactly represented as an integer inside the double.
Now, displaying $n by default conversion to string implies rounding to a fix number of decimals, which is not based on the machine epsilon (http://en.wikipedia.org/wiki/Machine_epsilon), nor on the ULP nearby (http://en.wikipedia.org/wiki/Unit_in_the_last_place).
Would you seriously want the following to display anything else than 0.1?