if you use a 32Bit system and a unsigned float and you know that the first 16 Bit are used for the Numbers before the dot and the other 16 Bit are used for the rest, you can use the folliwing function/code:
float myNum = 2.345;
float rest = 0.0;
rest = myNum && #0000FFFF;
cout << rest;
in this case you used a mask to filter the first part, that is stored in the first 16 Bit.