The standard libraries that support programming languages perform many common tasks.  C compilers ship with the libraries that include functions for mathematical calculations, generation of random events and manipulation and analysis of character data.  To access any function within any library we simply include the appropriate header file for that libray and call the function in our source code. 

This chapter introduces some of the more common functions in the libraries that ship with C compilers.  The GNU Documentation includes a comprehensive description of each function in each library. 

Mathematical Functions

The mathematics related libraries that contain the more common mathematical functions are:

• stdlib - the standard library
• math - the math library

Standard Library

The header file <stdlib.h> contains prototypes for the functions that perform the more general mathematical calculations.  These calculations include absolute values of integers and random number generation. 

Integer Absolute Value

abs(), labs(), llabs() return the absolute value of the argument.  Their prototypes are

 int abs(int);
 long labs(long);
 long long llabs(long long);

The following program produces the output on the right

 #include <stdlib.h>
 #include <stdio.h>

 int main(void)
 {
         int x = -12;
         long y = -24L;

         printf("|%d| is %d\n", x, abs(x)); 
         printf("|%ld| is %ld\n", y, labs(y)); 
         return 0;
 }

  

  
  
  
  
  
  
 |-12| is 12
 |-24| is 24
  
 

Random Numbers

rand() returns a pseudo-random integer in the range 0 to RAND_MAX.  RAND_MAX is implementation dependent but no less than 32767.  The prototype for rand() is

 int rand(void);

The following program outputs the same set of 10 pseudo-random integers for each successive run:

 #include <stdlib.h>
 #include <stdio.h>

 int main(void)
 {
         int i;

         for (i = 0; i < 10 ; i++)
                 printf("Random number %d is %d\n", i+1, rand()); 
         return 0;
 }

The following program outputs the same set of 10 pseudo-random integers between 6 and 100 inclusive:

 #include <stdlib.h>
 #include <stdio.h>

 int main(void)
 {
         int i, n, a = 6, b = 100;

         for (i = 0; i < 10 ; i++) {
                 n = a + rand() % (b + 1 - a);
                 printf("Random number %d is %d\n", i+1, n); 
         }
         return 0;
 }

The following program outputs the same set of 10 pseudo-random floating-point numbers between 3.0 and 100.0 inclusive:

 #include <stdlib.h>
 #include <stdio.h>

 int main(void)
 {
         int i;
         double x, a = 3.0, b = 100.0;

         for (i = 0; i < 10 ; i++) {
                 x = a + ((double) rand() / RAND_MAX * (b - a)); 
                 printf("Random number %d is %.2lf\n", i+1, x); 
         }
         return 0;
 }

rand() generates the same set of random numbers for every run of its host application.  This is very useful during the debugging stage.  To generate a different set of random numbers for every run we add a call to the function srand().  srand() sets the seed for the random number generator.  The prototype for srand() is

 int srand(unsigned seed);

unsigned is a type that only holds non-negative integer values.  We call srand() with time(NULL) (see below) as the argument once before the first call to rand(), typically at the start of our program.  This provides a unique seed for each run and hence different pseudo-random numbers.

The following program outputs a different set of 10 pseudo-random numbers with every run:

 #include <stdlib.h>
 #include <stdio.h>
 #include <time.h> // prototype for time(NULL)

 int main(void)
 {
         int i;

         srand(time(NULL));
         for (i = 0; i < 10 ; i++)
                 printf("Random number %d is %d\n", i+1, rand()); 
         return 0;
 }

math

The math library contains many functions that perform mathematical calculations.  Their prototypes are listed in <math.h>.  To compile a program that uses one of these function with the gcc compiler, we add the -lm option to the command line

 gcc myProgram.c -lm

Floating-Point Absolute Value

fabs(), fabsf(), fabsl() return the absolute value of the argument.  Their prototypes are

 double fabs(double);
 float fabsf(float);
 long double fabsl(long double);

The following program produces the output on the right

 #include <math.h>
 #include <stdio.h>

 int main(void)
 {
         float w = -12.5;
         double x = -12.5;

         printf("|%f| is %f\n", w, fabsf(w)); 
         printf("|%lf| is %lf\n", x, fabs(x)); 
         return 0;
 }

  
  
  

  
  
  
  
 |-12.500000| is 12.500000
 |-12.500000| is 12.500000
  
 

Floor

floor(), floorf(), floorl() return the largest integer value not greater than the argument.  Their prototypes are

 double floor(double);
 float floorf(float);
 long double floorl(long double);

For example, floor(16.3) returns a value of 16.0.

Ceiling

ceil(), ceilf(), ceill() return the smallest integer value not less than the argument.  Their prototypes are

 double ceil(double);
 float ceilf(float);
 long double ceill(long double);

For example, ceil(16.3) has a value of 17.0.

Rounding

round(), roundf(), roundl() return the integer value closest to the argument.  Their prototypes are

 double round(double);
 float roundf(float);
 long double roundl(long double);

For example, round(16.3) has a value of 16.0, while round(-16.3) returns a value of -16.0.

Truncating

trunc(), truncf(), truncl() return the integer part of the argument.  Their prototypes are

 double trunc(double);
 float truncf(float);
 long double truncl(long double);

For example, trunc(16.7) has a value of 16.0, while trunc(-16.7) returns a value of -16.0.

Square Root

sqrt(), sqrtf(), sqrtl() return the square root of the argument.  Their prototypes are

 double sqrt(double);
 float sqrtf(float);
 long double sqrtl(long double);

For example, sqrt(16.0) returns a value of 4.0.

Powers

pow(), powf(), powl() return the result of the first argument raised to the power of the second argument.  Their prototypes are

 double pow(double base, double exponent);
 float powf(float base, float exponent);
 long double powl(long double base, long double exponent);

The following program produces the output on the right

 #include <math.h>
 #include <stdio.h>

 int main(void)
 {
         double base = 12.5;

         printf("%lf^3 is %lf\n", base,
          pow(base,3)); 
         return 0;
 }

  
  
  

  
  
  
  
 12.500000^3 is 1953.125000 
  
 

This set of functions was designed for floating-point arguments and each one is computationally intensive.  Avoid using them for simpler integer arguments. 

Logarithms

log(), logf(), logl() return the natural logarithm of the argument.  Their prototypes are

 double log(double);
 float logf(float);
 long double logl(long double);

For example, log(2.718281828459045) returns a value of 1.0.

Powers of e

exp(), expf(), expl() return the natural anti-logarithm of the argument.  Their prototypes are

 double exp(double);
 float expf(float);
 long double expl(long double);

For example, exp(1.0) returns a value of 2.718281828459045.

Time Functions

The time library contains functions that return timing data.  Their prototypes are listed in <time.h>.

Calendar Time

time

time(NULL) returns the current calendar time.  The prototype is

 time_t time(time_t *);

time_t is a type that it is sufficiently large to hold time values - for example, unsigned long. 

difftime

difftime() returns the difference in seconds between two calendar time arguments.  The prototype is

 double difftime(time_t, time_t);

time_t is a type that it is sufficiently large to hold time values - for example, unsigned long. 

The following program returns the time in seconds taken to execute the central iteration

 #include <time.h>
 #include <stdio.h>
 #define NITER 1000000000

 int main(void)
 {
         double x;
         int i, j, k;
         time_t t0, t1;

         x = 1;
         t0 = time(NULL);
         for (i = 0; i < NITER; i++)
                 x = x * 1.0000000001;
         t1 = time(NULL);
         printf("Elapsed time is %.1lf secs\n", 
          difftime(t1, t0)); 
         printf("Value of x is %.10lf\n", x);

         return 0;
 }

  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
 Elapsed time is 40.0 secs
 Value of x is 1.1051709272 
  
  
 

Process Time

clock

clock() returns the approximate process time.  The prototype is

 clock_t clock(void);

clock_t is a type that holds time values - for example, unsigned long.  The value is in units of CLOCKS_PER_SEC.  We divide by these units to obtain the process time in seconds.

The following program returns the process time in seconds taken to execute the central iteration

 #include <time.h>
 #include <stdio.h>
 #define NITER 400

 int main(void)
 {
         double x;
         int i, j, k;
         time_t t0, t1;
         clock_t c0, c1;

         x = 1;
         t0 = time(NULL);
         c0 = clock();
         for (i = 0; i < NITER; i++)
             for (j = 0; j < NITER; j++)
                 for (k = 0; k < NITER; k++)
                     x = x * 1.0000000001;
         t1 = time(NULL);
         c1 = clock();
         printf("Elapsed time is %.1lf secs\n",
                difftime(t1, t0)); 
         printf("Process time is %.3lf secs\n",
                (double)(c1-c0)/CLOCKS_PER_SEC); 
         printf("Value of x is %.10lf\n", x);

         return 0;
 }

  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
 Elapsed time is 1.0 secs
 Process time is 0.040 secs
 Value of x is 1.0001000050 
  
  
  
  
 

The cast to a double forces floating-point division.

Character

The ctype library contains functions for character manipulation and analysis.  Their prototypes are listed in <ctype.h>. 

Manipulation

tolower

tolower() returns the lower case of the character received, if possible; otherwise the character received unchanged.  The prototype is

 int tolower(int);

For example, tolower('D') returns 'd', while tolower(';') returns ';'.

toupper

toupper() returns the upper case of the character received, if possible; otherwise the character received unchanged.  The prototype is

 int toupper(int);

For example, toupper('d') returns 'D', while toupper(';') returns ';'.

Analysis

The character analysis functions return 'true' or 'false'.  C represents 'false' by the value 0 and 'true' by any other value. 

islower

islower() returns a true value if the character received is lower case, a false value otherwise.  The prototype is

 int islower(int);

For example, islower('d') returns a true value, while islower('E') returns a false value.

isupper

isupper() returns a true value if the character received is upper case, a false value otherwise.  The prototype is

 int isupper(int);

For example, isupper('d') returns a false value, while isupper('E') returns a true value.

isalpha

isalpha() returns a true value if the character received is alphabetic, a false value otherwise.  The prototype is

 int isalpha(int);

For example, isalpha('d') returns true, while isalpha('6') returns false.

isdigit

isdigit() returns a true value if the character received is a digit, a false value otherwise.  The prototype is

 int isdigit(int);

For example, isdigit('d') returns false, while isdigit('6') returns true.

isspace

isspace() returns a true value if the character received is a whitespace character, a false value otherwise.  The prototype is

 int isspace(int);

For example, isspace('d') returns false, while isspace(' ') returns true.  Whitespace characters are ' ', '\t', '\n', '\v', and '\f'.

isblank

isblank() returns a true value if the character received is a space or tab, a false value otherwise.  The prototype is

 int isblank(int);

For example, isblank('\t') returns true, while isblank('\n') returns false.

Exercises

• Complete the Workshop on Library Functions