Option B and D are the correct answer.
Since the variables I and j are floats, resultant will be float type too. So we have to use float
or primitive type which can hold float, such a primitive type is double, it has wider range
and also can hold floating point numbers, hence we can use double or float for the blank.
As explained above options B and D are correct.
long and int can't be used with floating point numbers so option A is incorrect.
Option E is incorrect as it have smaller range and also can't be used with floating point
numbers.
hnpsy/docs.oracle.com/javase/tutorial/java/javaOO/variables.html