float holds value of -NAN

[Wings]

hi,
A float I'm reading periodically holds a value of -NAN, I was wondering if anyone has run into this problem before and can tell me what it means and a valid way of checking for it? Or a way to check that a float holds a valid float value would be helpful to.

Thanks

 

[xwb]

It just means that it is bigger than the biggest number allowable. With floats, not all the bit patterns are used. Two of the patterns are illegal and they result in Nan or -Nan. Say grandma gives you Nan. What you could do is

unsigned char* cgran = (unsigned char*) &grandma;
printf ("%02x %02x %02x %02x\n", cgran[0], cgran[1], cgran[2], cgran[3]);

This will give you the hex representation of the sequence that is giving you the Nans. It is normally quite a big number. Could there be an alignment problem with your data?