Type casting integer to float ?

RE: Type casting integer to float ?

Felix Tam — Sat, 06 Feb 2021 10:07:37 GMT

Float in c is IEEE754 format. 1st bit represent sign, the next 8 bits represent expo and the rest 23 bits

represent fraction. The uint32_t you try is 2248 that the expo is 0. Try this

float x = 1.1;

uint32_t int_value = *(uint32_t *)&x;

//int_value is 1066192077

float y = *(float *) &int_value;

Or use memcpy, the effect is the same.

RE: Type casting integer to float ?

RixtronixLAB — Sat, 06 Feb 2021 07:54:28 GMT

From terminal :

<info> app: Float number: 0.000





nfo> app: Float number V2:

Code :

	uint32_t int_value;
	float float_val;
	
       
    //float_val = *(float *)&int_value;	
    memcpy(&float_val, &int_value, sizeof(int_value));
	  NRF_LOG_INFO("Float number: "NRF_LOG_FLOAT_MARKER"\n\r",NRF_LOG_FLOAT(float_val));
	  NRF_LOG_INFO("Float number V2: %.2f",float_val);

RE: Type casting integer to float ?

RixtronixLAB — Sat, 06 Feb 2021 06:23:17 GMT

Snipped :

	unsigned int dat = 2248;
	
	

      
    float float_val = *(float *)&dat;	
    
	  NRF_LOG_INFO("Float number: "NRF_LOG_FLOAT_MARKER"\n\r",NRF_LOG_FLOAT(float_val));
	  NRF_LOG_INFO("Float number V2: %.2f",float_val);

I got 0.00 for float_val ??

RE: Type casting integer to float ?

RixtronixLAB — Sat, 06 Feb 2021 03:30:56 GMT

then NRF_LOG_INFO("Float number V2: %.2f",float_val); ?

RE: Type casting integer to float ?

RixtronixLAB — Sat, 06 Feb 2021 02:58:34 GMT

Hi Einar,

Thanks for the reply,

I will fix it and keep posted, thanks again

RE: Type casting integer to float ?

Felix Tam — Fri, 05 Feb 2021 20:40:47 GMT

uint32_t int_value;

float float_val;

float_val = *(float *)&int_value;

or memcpy(&float_val, &int_val, size_of(int_value);

RE: Type casting integer to float ?

Einar Thorsrud — Mon, 01 Feb 2021 12:55:02 GMT

Hi,

There is a mistake in your code. your float_val variable is a float (32 bit), but you cast dat to a double (64 bit). So you should fix the fast so that you cast it to a float instead.