0

I know that the results are same, but the first loop is faster than the second loop. Why is that?

int[][] values = new int[10][20]; int sum = 0; int count = 0; for (int j = 0; j < 20; j++) { for (int i = 0; i < 10; i++) { count++; values[i][j] = count; sum += values[i][j]; } } System.out.println(sum); 
int[][] values = new int[10][20]; int sum = 0; int count = 0; for (int i = 0; i < 10; i++) { for (int j = 0; j < 20; j++) { count++; values[i][j] = count; sum += values[i][j]; } } System.out.println(sum);
Improve this question
3
  • What is your doubt? What do you want to know? Commented May 11, 2015 at 5:49
  • the @joshstrike answer is right! Commented May 11, 2015 at 5:52
  • I tried your codes, both are giving the same output: 20100. So nothing is wrong Commented May 11, 2015 at 6:00

2 Answers 2

Reset to default
3

This is a loop interchange.

While I can't exactly observe the same results as your machine (mine is quite beefy with an Intel i7-4770K, and the code is in a very simple class I use for Stack Overflow questions), in general this sort of loop has the propensity to be more performant.

The primary reason that it's more performant is due to the locality of the elements it's iterating over. It is more efficient to iterate over contiguous memory than it is non-contiguous memory.

Let's consider a typical array:

int[] anArray = new int[5];

In memory, Java will reserve* five contiguous blocks of that are capable of holding an int.

int[5] -> [][][][][]

*: Or at least try to.

For a two-dimensional array, Java will do the same operation, with the catch that instead of it appearing like a grid in memory (as C does), you will have an array that points to arrays.

int[2][3] -> [ ] [ ] | | v v [ ] [ ] [ ] [ ] [ ] [ ]

In terms of memory layout, the second-tier arrays (that is, those of length 20) are likely closer in memory to each other than they are with the first-tier arrays (that is, those of length 10). It's more efficient to iterate over contiguous memory than it is non-contiguous memory, which is why you're noticing a speed improvement.

Improve this answer
Sign up to request clarification or add additional context in comments.

1 Comment

Yes, and that explains why the second way of looping should be faster than the first one. But the OP claims the opposite.
0

Because you forgot to switch [i][j] in values when you push them into that array.

Improve this answer

6 Comments

switching i and j would give an ArrayIndexOutOfBoundsException.
values[i][j] is probably correct in the second version and was wrong in the first, but in any event it may need to be tested or defined first, point being, that's why his two pieces of code are producing two different sums.
They both produce the same sum - 20100.
argh, what is the point. Okay so like, the title of this question is "sum of array calculation difference" and during the guy's repeated edits of it I'm pretty sure he wiped out the original question and part of the offending code. This site wears me down.
If the question is not clear, there's no point in answering it until it is clarified.
|

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.