FLOAT(a,p) is an arithmetic function that outputs the expression a (a number) in a format of the data type FLOAT(p) (p is an unsigned integer). It is rounded to p places if necessary.