You could print the integer part and then do a fixed number of decimal places.
Something like,
Code:
float x=3.1416;
myint=(int)x;
//send myint to LCD
//send decimal point to LCD
x-=myint; //keep only decimal part
for(i=0;i<4;i++){
x*=10;
myint=(int)x;
x-=myint;
//send myint to LCD
}