division not returnig float

MorgoZ · July 2015

Hi,

I don't know why this operation returns an integer value and not a float:

DEFINE_FUNCTION char[10] DbsIORealToFake(char cTipoIO, float real)
{
    stack_var float fAux
        integer nAux, nAuxTp, nAuxFake
    
    nAuxTp = MAX_LEVEL_TP - MIN_LEVEL_TP
    
    if(real > MAX_LEVEL_TP)
    real = MAX_LEVEL_TP
    if(real < MIN_LEVEL_TP)
    real = MIN_LEVEL_TP       
    
    fAux = (real - MIN_LEVEL_TP) / nAuxTp
    send_string 5001:0:0,"'fAux ',ftoa(fAux)"
}

The last line returns "0" when it should be "0.3". Does enyone can explain why is it being rounded to an integer?

Note: "real" comes from a level value

Thanks!

ErikMeyer · July 2015

I believe the compiler will see that the value of real is an integer (since it is from a level) and recasts during the math operation since the other operands are also integers. So you essentially have float = (integer - integer) / integer. Thus the math operation returns an integer. To get a float value back, I believe you will need to define nAuxTP as a float and create a new variable fMin and assign MIN_LEVEL_TP as follows:

float fAuxTP
float fMin

fMIN = MIN_LEVEL_TP

...

fAux = (real - fMin) / fAuxTP

Now you have float = (float - float) / float which should return a float type. I have also heard, but not tried, that simply adding and then subtracting 0.1 to the equation will cause the compiler to return a float as follows:

fAux = (real - MIN_LEVEL_TP + 0.1 - 0.1) / nAuxTp

NZRob · July 2015

Just wanted to check if you are talking about when you say return - if you are talking about what is returned from the function and not the send_string then you are missing the RETURN command

e.g.
At the end of the function you need:

RETURN ftoa(fAux)

Then char[10] will be something else, it will be nothing

MorgoZ · July 2015

Ok, i had to define at least one variable at the division operands as a float and then the return type was also float, as EricMeyer said.

Thank you both!

division not returnig float

Comments