a casting problem in the textbook — is it 0.05 or 0.5?

Question

double cost = 10.95;
int numDollars = (int) cost; //sets numDollars to 10

If your intent was to round cost to the nearest dollar, you needed to write

int numDollars = (int)(cost+0.5); //numDollars has value 11

Above is what's written in my AP computer science A Barron's book. I'm quite new to Java and all, but I was just wondering if the 0.5's should be replaced with 0.05's. If the book is right though, could someone help me understand why?

Answer 1

The example uses 10.95, but the solution needs to work for all values.

Consider a different example using 10.51
What should the right answer be? 11

But how do you get there?

10.51 + 0.50 = 11.01, truncated by typecast to 11

If we tried to use your suggestion, we would get:

10.51 + 0.05 == 10.56, truncated by typecast to 10 Wrong Answer