A simple yet a valid question...

Hello,
Please tell me why can't i use a float variable in structure??

for eg..

Code:

---

struct Point
{
    float x;
    float y;
};

---

Now when i use this structure and take variable input using scanf, it gives error " Floating Point Format Not Linked"
and then without taking any input it displays message "Abnormal Program Termination" and exits.

Please tell why is this happening and also it's solution

Quote:

---

Please tell me why can't i use a float variable in structure??

---

But you can.

Quote:

---

Now when i use this structure and take variable input using scanf, it gives error " Floating Point Format Not Linked"
and then without taking any input it displays message "Abnormal Program Termination" and exits.

---

Perhaps you are using it wrongly. Post the smallest and simplest program that demonstrates the error.

Code:

---

struct Point
{
    float x;
    float y;
};

int main() {
    struct Point p;
    scanf("%f",&(p.x));
}

---

This should compile. Maybe you are missing the struct keyword in front of Point p;? Maybe you are doing scanf("%f",p)?

My bet would be that you are using Turbo C or some other ancient compiler. In old times, processors were not always able to use floating point, and the additional library for floating point in C was quite large, the compiler would have one set of "integer library" and another "floating point library".

You need to read the help for your compiler to see what you should enable to get your floating point included.

However, the other question would be why you are not using a more modern compiler.

--
Mats

Perhaps i didn't convey my problem correctly. Actually problem only occurs when you define an array of struture which contains only float values. Here's a small example for that

Code:

---

#include <stdio.h>

struct points
{
  float x;
  float y;
};

int main(void)
{
  struct points p[20];
  int vertices,i;

 printf("\n\nEnter The Number Of Vertices :");
 scanf("&#37;d",&vertices);
 for(i=0;i<vertices;i++)
 {
        printf("\n\nEnter The %d th Point :",i+1);
        scanf("%f %f",&p[i].x,&p[i].y);
 }

 return 0;
}

---

Yeah matsp, you are correct ( as usual). I am using ancient Turbo C++ 3.0 compiler

I compiled (MinGW port of gcc 3.4.5) and ran your program with no problems.

What compiler are you using?

EDIT:
Ah.

Some googling comes up with this:
http://www.jeffc.org/msdos/dos-faq.html#Q0304

--
Mats

Thanks Mats!! :)
Now it's working fine

Any reason you can't switch compiler?

Oh I remember the days you had to force the linker to include the math library. Even switches on whether to in-line 387 code or call an emulation library, or thirdly, to include auto-detect of the target machine if you're not sure if it had floating point support.

Life was simpler back then. (Well, yeah, it was, because I understood why things were the way the were.

The reason is in my college it's the same compiler since last 10 years and i brought the compiler from my college and therefore i use that compiler!when i finish college i will plan to use a better compiler
As such my job is done on this compiler too

> Oh I remember the days you had to force the linker to include the math library
Those days haven't gone anywhere :)

Quote:

---

Originally Posted by chottachatri

The reason is in my college it's the same compiler since last 10 years and i brought the compiler from my college and therefore i use that compiler!when i finish college i will plan to use a better compiler
As such my job is done on this compiler too

---

It doesn't need to be.
They scammed you when they sold you that compiler. They took money for something that can be found for free, and it isn't even standards compliant to boot.
My suggestion: dump it and get a better one. Right now.
Doesn't matter if you bought it or not.