Decimal Numbers

An int holds only whole numbers. But a price can have cents, and a height can be 1.68 m. For numbers with a decimal part, C has the type double:

double price = 12.5;

In C, decimal numbers are always written with a decimal point: 12.5, -0.25, 3.0.

To print a double, use %f instead of %d (f as in floating point, the way computers store decimal numbers). %f always prints six decimal places. Usually, you want fewer, so write their number between % and f, after a dot: %.2f prints two decimal places, %.1f one. The number is rounded:

double price = 12.5;
double pi = 3.14159;
printf("%f\n", price);
printf("%.2f\n", price);
printf("%.1f\n", price);
printf("%.2f\n", pi);

This prints:

12.500000
12.50
12.5
3.14

For prices, %.2f is just right: two decimal places, as on a receipt.

The format specifier must fit the value: %d for an int, %f for a double. printf itself can't check what it gets; it trusts the format string. On another computer, a wrong specifier could print a nonsense number, so on this site, it's a compile error:

#include <stdio.h>

int main(void) {
    double price = 12.5;
    printf("Price: %d\n", price);
    return 0;
}

The compiler reports:

main.c:5:27: error: format specifies type 'int' but the argument has type 'double' [-Werror,-Wformat]

The format asks for an int, but the value (the argument) is a double. Below the first line, the compiler even suggests %f. It works the other way round, too: %f with an int ends with format specifies type 'double' but the argument has type 'int'.

More typical mistakes:

• A decimal comma: in many countries, people write 12,5. In C, double price = 12,5; ends with the error expected identifier or '(', because the comma means something else in C.
• A decimal number in an int: int price = 12.5; compiles, but only 12 is stored. The decimal part is thrown away.