Why we use %g in C

To read an integer we use either %d and %i format strings, but to read a float we use %e, %f, and  %g format strings of type double

The %f format string

To begin with, suppose, we have a float variable a=32 and, we would like to print variable a with %f format string

 float a=32.0f;
    printf("%f",a); //result 32.000000

Guess what the output will be?

The answer is 32.000000

But wait!

Why does the output have 6 fractional digits?

as can be seen, The %f format string will always display the floating point variable in a fixed decimal format which means it will always have a fraction part which is by default six (.000000)

with this intention,32 will become 32.000000,

The %e format string

The %e format string will always display the floating point variable in exponential format

float a=32.0f;
    printf("%e",a);// the result is 3.200000e+001

As I have shown, if we use %e we get the answer as 3.200000

Here, the 200000 is the default 6-fraction digits

But why

Because we right-shifted the decimal point by one digit right

Number	exponential notation	scientific notation
1000	1e3	1×103
1000000	1	1x
1000000000	1	1x

The %g format string

The %g format string can display the floating point variable in either exponential format or fixed decimal format  

But, how %g decide to display exponential format or fixed decimal format?

At this point, It all depends upon the floating-point variable size

In short, it switches  to exponential format when a floating point variable exponent is less than -4(that is, 0.0004=4x=4) or greater than the precision(which is by default 6) such as[1234567]

Trailing zeros and trailing decimal point are not printed

float a=32.0f;
    printf("%g",a); // the result is 32

Here, the % g format string is not displaying the trailing zeros of 32

However, the %f format string can display the trailing zeros of 32.000000

Therefore, the %f format string can display 32.23 as 32.230000 but the %g format string can display 32.23 as 32.23(the trailing zeros are neglected)

Furthermore, you can print the trailing zeros and trailing decimal points with %g format string by adding the #flag

float a=32.0f;
    printf("%#g",a); // the result is 32.0000

Later, if the floating variable is too large, or too small then, at that instant, the %g format string switches to the exponential format

so, that variable adjust itself with fewer characters

float a=321224587.23f;
    printf("%g",a); // the result is 3.21225e+008

To read and print the double value with %g format string, use l(ell) with scanf for %g format string(%lg), but, the l(ell) will not have any effect on printf function with %g format string.

double a=32.23;
   double b;
printf("enter the value of B\n");
    scanf("%lg",&b);
    printf("VALUE of a= %g\n",a); // the result is 32.23
    printf("VALUE of b= %g\n",b);

In C99 to read and write the long double use the letter L in front of the %g format string

long double a=32.23;
   scanf("%Lf",&a);
    printf("VALUE of a= %Lg\n",a);

Instead, if you use the C89 version of C then the effect of using conversion specifications %g and %G is undefined but, it is legal in C99