issues with expr

[vsdpsingh]

Hi,

tclsh
% expr 0.2+0.2
0.4
% expr 0.2+0.1
0.30000000000000004
%

Can anyone please explain me the logic, that if i add 0.2 with 0.2, it will give 0.4
but, if i add 0.2 with 0.1, it will give 0.30000000000000004 and not 0.3


Am i missing some thing.
thanks in advance.

 

[mikrom]

It's because of floating point arithmetic
Use format statement

% format "%5.3f" [expr 0.2+0.2]
0.400

 

[feherke]

Hi

That is because the internal representation of floating point values used by computers. That behavior can be observed in many other languages :

master # cat precision.java
public class precision {
  public static void main(String[] a)
  {
    System.out.println(0.2+0.2);
    System.out.println(0.2+0.1);
  }
}

master # javac precision.java

master # java precision
0.4
0.30000000000000004

>>> 0.2+0.2
0.4
>>> 0.2+0.1
0.30000000000000004

>>> 0.2+0.2
0.4
>>> 0.2+0.1
0.30000000000000004

irb(main):001:0> 0.2+0.2
=> 0.4
irb(main):002:0> 0.2+0.1
=> 0.30000000000000004

The difference is mostly only in whether the language designers decided to beautify the output of floating point values, or not. But even if the output is beautified, behind the scenes in most cases the same internal representation is used, so you must be aware of it when comparing floating point values :

 > =0.2+0.2
0.4
> =0.2+0.1
0.3
> =0.2+0.2==0.4
true
> =0.2+0.1==0.3
false

  DB<1> print 0.2+0.2
0.4
  DB<2> print 0.2+0.1
0.3
  DB<3> print 0.2+0.2==0.4 ? 'equal' : 'not'
equal
  DB<4> print 0.2+0.1==0.3 ? 'equal' : 'not'
not

Feherke.
feherke.github.io

 

[vsdpsingh]

Thanks for reply. I will add check for the same.