I have tried values 0.1,0.2,0.3,0.35,0.3000025. None seem to print the word "If", they print "Else".Code:#include<stdio.h> #include<conio.h> void main() { float x=0.1; if(x==0.1) printf("If"); else printf("Else"); getch(); }
Could somebody help me with the reason for the same.
0.1 can be stored internally as 0.099999
or 0.100001 for example
you can never be sure...
so if you really need to compare floats do something like
where delta is small enough for considering to float numbers as the sameCode:if(abs(x-y)<delta)
All problems in computer science can be solved by another level of indirection,
except for the problem of too many layers of indirection.
– David J. Wheeler
Vart i do agree with the part of 0.1 being stored that way, but why is it so and even if it is so. If it stores 0.1 as 0.099999 or 0.1000001 then why does it not compare it also that way?
The problems began with void main.
Then this happened - http://c-faq.com/fp/strangefp.html
If you dance barefoot on the broken glass of undefined behaviour, you've got to expect the occasional cut.
If at first you don't succeed, try writing your phone number on the exam paper.
'cause in the statement x == 0.1
x is a float and 0.1 is double
so they can be stored as 0.999999 and 0.999999999999999
then both converted to double will give different values...
for example...
as I said - you never know how the float value is stored internally
All problems in computer science can be solved by another level of indirection,
except for the problem of too many layers of indirection.
– David J. Wheeler
> if(x==0.1)
Because the left hand side is a float, and the right hand side is a double, and in the float to double conversion, you get a slightly different approx of 0.1, one which is different.
if ( x == 0.1f )
Might just work, but as stated, you should never compare floats for equality.
If you dance barefoot on the broken glass of undefined behaviour, you've got to expect the occasional cut.
If at first you don't succeed, try writing your phone number on the exam paper.
If i do this it returns the value of x as 0.100000.Code:else printf("%f,x"); getch(); }
As per ur link salem, i understand that the processor's registers and the memory store values differently. So is there no way perfect way for equating floats?
Vart do u mean to say that the rvalue of the if statement is by default a "double"?
a constant is by default a double
All problems in computer science can be solved by another level of indirection,
except for the problem of too many layers of indirection.
– David J. Wheeler
> So is there no way perfect way for equating floats?
Floats and doubles are always approximations.
Find the David Goldberg, ``What Every Computer Scientist Should Know about Floating-Point Arithmetic''
If you dance barefoot on the broken glass of undefined behaviour, you've got to expect the occasional cut.
If at first you don't succeed, try writing your phone number on the exam paper.
Thx guys u actually genuinely solved a query which i was banging my head for, I am from India and we have teachers with bookish knowledge so wouldnt expect an answer from them.
