Using floats with sprintf() in embedded C

Question

Guys, I want to know if float variables can be used in sprintf() function.

Like, if we write:

sprintf(str,\"adc_read = %d \         


        

Answer 1

use the %f modifier:

sprintf (str, "adc_read = %f\n", adc_read);

For instance:

#include <stdio.h>

int main (void) 
{
    float x = 2.5;
    char y[200];

    sprintf(y, "x = %f\n", x);
    printf(y);
    return 0;
}

Yields this:

x = 2.500000

Answer 2

Similar to paxdiablo above. This code, inserted in a wider app, works fine with STM32 NUCLEO-F446RE.

#include <stdio.h>
#include <math.h>
#include <string.h>
void IntegFract(char *pcIntegStr, char *pcFractStr, double dbValue, int iPrecis);

main()
{
   char acIntegStr[9], acFractStr[9], char counter_buff[30];
   double seconds_passed = 123.0567;
   IntegFract(acIntegStr, acFractStr, seconds_passed, 3);
   sprintf(counter_buff, "Time: %s.%s Sec", acIntegStr, acFractStr);
}

void IntegFract(char *pcIntegStr, char *pcFractStr, double dbValue, int 
iPrecis)
{
    int iIntegValue = dbValue;
    int iFractValue = (dbValue - iIntegValue) * pow(10, iPrecis);
    itoa(iIntegValue, pcIntegStr, 10);
    itoa(iFractValue, pcFractStr, 10);
    size_t length = strlen(pcFractStr);
    char acTemp[9] = "";
    while (length < iPrecis)
    {
        strcat(acTemp, "0");
        length++;
    }
    strcat(acTemp, pcFractStr);
    strcpy(pcFractStr, acTemp);
}

counter_buff would contain 123.056 .

Answer 3

Isn't something like this really easier:

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

char str[10];
float adc_read = 678.0123;

dtostrf( adc_read, 3, 4, temp );
sprintf(str,"adc_read = %10s \n", temp);
printf(temp);

Answer 4

Yes, use %f

Answer 5

Don't do this; integers in C/C++ are always rounded down so there is no need to use the floor function.

char str[100]; 
int d1 = value;

Better to use

int d1 = (int)(floor(value));

Then you won't get rounding up of the integer part (68.9999999999999999 becomes 69.00..). 68.09999847 instead of 68.1 is difficult to avoid - any floating point format has limited precision.

Answer 6

Yes, and no. Despite what some other replies have said, the C compiler is required to perform conversions for sprintf(), and all other variadic functions, as follows:

• char => int
• short => int
• float => double

(and signed/unsigned variants of the above integral types)

It does this precisely because sprintf() (and the other print()-family functions) would be unusable without it. (Of course, they're pretty unusable as it is.)

But you cannot assume any other conversions, and your code will have undefined behaviour - read: crash! - if you do it.