the float data type is a 32bit value, if you convert it to a variable that uses only one byte like an unsigned char then you will loose all the decimals and you can only represent integers from 0 to 255 (0b00000000 to 0b11111111).
Assuming that the value can fit in 8bits and you don't mind truncating the decimals you can use (unsigned char)my_float_variable
Are the float values higher than zero or they can be negative too?