I need to convert a float, including negative floats, to a char array.
I have tried a lot of things and can't get anything to work. This is my last effort.
// val is a float
int val_int = (int) val; // compute the integer part of the float
float val_float = (abs(val) - abs(val_int))*100000;
int val_fra = (int)val_float;
sprintf (sz, "%d.%d", val_int, val_fra); //
When the float is -96.12345, I get -96.0
Any help would be appreciated.
Dana
Google for:
Arduino sprint float
You'll see an explanation and some imaginative work arounds.
Thanks.
I have tried all that I can find and when it does (work) it gets the decimal part wrong. I hard coded
-96.123465.
This, taken more or less from your code, on an arduino Uno :
void setup() {
Serial.begin( 115200 ) ;
float val = -96.12345 ; // test value
char sz[20] = {' '} ;
int val_int = (int) val; // compute the integer part of the float
float val_float = (abs(val) - abs(val_int)) * 100000;
int val_fra = (int)val_float;
sprintf (sz, "%d.%d", val_int, val_fra); //
Serial.println ( sz ) ;
}
void loop() {
}
gives this result:
-96.12345
Do you need more precision than that ?
Better use this:
sprintf (sz, "%d.%05d", val_int, val_fra); //
Otherwise you'll get "0.5" when the value is "0.00005" or "0.0005" or any other similar value.
(Better yet, use the well-tested "ftoa" (correction: "dtostrf") function that is included in the AVR libc.)
Better yet, use the well-tested "ftoa" function that is included in the AVR libc.
Can you provide a link for that. All that I see avr-libc 2.0.0 Standard C library for AVR-GCC is dtostrf() in stdlib.h.
Just use dtostrf(). It works just fine with floats
6v6gt:
This, taken more or less from your code, on an arduino Uno :void setup() {
Serial.begin( 115200 ) ;
float val = -96.12345 ; // test value
char sz[20] = {' '} ;
int val_int = (int) val; // compute the integer part of the float
float val_float = (abs(val) - abs(val_int)) * 100000;
int val_fra = (int)val_float;
sprintf (sz, "%d.%d", val_int, val_fra); //
Serial.println ( sz ) ;
}
void loop() {
}
gives this result: -96.12345 Do you need more precision than that ?
Serial.print will work, but I am not using that. I am using the sz in another function. That function works
because I tested it using strcpy and then sending that string to the function.
cattledog:
Can you provide a link for that. All that I see avr-libc 2.0.0 Standard C library for AVR-GCC is dtostrf() in stdlib.h.
Sorry, you're right. I meant dtostrf(). I was thinking it was called ftoa() because itoa() exists to convert an int to string.
TXEng:
Serial.print will work, but I am not using that. I am using the sz in another function. That function works
because I tested it using strcpy and then sending that string to the function.
In principle, it is always good to provide a minimal program which demonstrates the problem. It seems that your problems may then lie in the later processing if you have succeeded in correctly forming a character array based on the float number.
UKHeliBob:
Just use dtostrf(). It works just fine with floats
Thanks, that seems to work fine.
Dana