why doesnt this work!!!

[666kennedy]

this isnt working, i think it has something to do with the natural exponential, its a calculation of the sigmoid function. it basically should read

1/(1 + e^-((W1A*pix1)+(W2A*pix2)))

this is what ive coded

OUTA = 1/(1 + exp -((W1A * pix1) + (W2A * pix2)));

please help me.

[Ancient Dragon]

your code needs to call pow() to raise x to the power of y. Don't know what exp is in your code.

[666kennedy]

exp is meant to represent the natural exponential, is how it is represented in every other electronic form ive met, i just assumed c++ would be the same, how would i incorporate this pow() in the code provided

[Ancient Dragon]

It's just a simple function call. Read this:. The paramter [x] would be the value of e (whatever that is) and the paramter y is the calculation -((W1A * pix1) + (W2A * pix2)));

[666kennedy]

thank you, helped lots, i just assumed that there was a symbol for the natural exponential in c++.

i am also creating random numbers for each weight.

float W1A =  -1 + rand() / ( RAND_MAX / ( 1 + 1 ) + 0.1 );
    float W1B =  -1 + rand() / ( RAND_MAX / ( 1 + 1 ) + 0.1 );
    float W2A =  -1 + rand() / ( RAND_MAX / ( 1 + 1 ) + 0.1 );
    float W2B =  -1 + rand() / ( RAND_MAX / ( 1 + 1 ) + 0.1 );

but every time i compile and run the program the same numbers come up. how can i make it create a new set of numbers each time.

[Ancient Dragon]

you need to seed the random number generator. Most people see it with the current clock time of the computer

#include <ctime>

...
srand( time(0) );

Depending on your compiler you may need to typecast the return value of time() to size_t.

[666kennedy]

i dont quite understand, but thats purely down to my knowledge of c++ i really am a beginner so sorry and i thank you for your patience, basically my system runs through and stops, i exit the console and if i want to run through it again i compile and run it again. so for everytime it runs through i want it to produce a different set of numbers. is that what this will do, how would i impliment it, what would x be in time(x)?

thanks again for your help

[Ancient Dragon]

I already showed you the code to do it. Just paste that line after main().

#include <ctime>
// other includes here

int main()
{
   srand( time(0) );
   // other code here
}

The time() function gets the current system time in seconds since 1970. So every time you run the program time() will return something different.

[666kennedy]

thanks bud, thats perfect, i dont think ill use it til the end so i can use the numbers to calculate the outcomes by hand too. anyway thanks so much. ill be asking another question soon, but not on this thread, as it is a different topic