In previous post, we compared LinkedList and ArrayList in addition operations using JMH.

In this post, we are going to compare ArrayList and LinkedList performance using deletion operations.

TRY IT YOURSELF: You can find the source code of this post here.

Java Collections List Series

• Part 1: Java Collections: List
• Part 2: ArrayList vs LinkedList: Addition
• Part 3: ArrayList vs LinkedList: Deletion (You are here)
• Part 4: ArrayList vs LinkedList: Sort, Get and Iteration

Environment

• JDK 11.0.3, OpenJDK 64-Bit Server VM, 11.0.3+7-Ubuntu-1ubuntu218.04.1
• Ubuntu 18.04
• Memory 15,6 GiB
• Processor Intel® Core™ i7-7500U CPU @ 2.70GHz × 4
• Graphics Intel® HD Graphics 620 (Kaby Lake GT2)
• OS 64-bit
• Disk 387,9 GB
• Intellj IDE Community 2018.3

Approach

• We are going to use JMH for the micro benchmark.
• We are going to use a list of 10.000.000 of elements, generated randomly, and we are going to calculate the average response time in 10 iterations.
• As JMH doesn’t allow sharing data between benchmarks, we are going to create a file with the list of elements, and read from it by each benchmark. In that way, we guarantee we are measuring the same scenario.

Deleting from the End – remove(size - 1)

Now, let’s see the code for measuring how ArrayList and LinkedList behave when we need to remove from the end:

  @Test
  public void remove_end()
          throws RunnerException, IOException {
    Utils.populateToFile();

    Options opt = new OptionsBuilder()
            .include(
                    ArrayListVsLinkedListDeletionEndTest.class
                            .getSimpleName())
            .detectJvmArgs()
            .build();

    Collection<RunResult> runResults = new Runner(opt)
            .run();

    Result arrayListResult = Utils
            .find("arrayList",
                    runResults);
    Result linkedListResult = Utils
            .find("linkedList",
                    runResults);

    assertThat(linkedListResult.getScore())
            .isLessThan(arrayListResult.getScore());
  }

  @Benchmark
  @Measurement(iterations = Utils.amountIterations, batchSize = 1, time = 1)
  @BenchmarkMode(Mode.SingleShotTime)
  @Fork(1)
  @Warmup(iterations = 10)
  @OutputTimeUnit(TimeUnit.MILLISECONDS)
  public ArrayList<Integer> arrayList(
          ArrayListExecutionPlan executionPlan) {
    ArrayList<Integer> list = executionPlan.arrayList;

    list.remove(list.size() - 1);

    return list;
  }

  @Benchmark
  @Measurement(iterations = Utils.amountIterations, batchSize = 1, time = 1)
  @BenchmarkMode(Mode.SingleShotTime)
  @Fork(1)
  @Warmup(iterations = 10)
  @OutputTimeUnit(TimeUnit.MILLISECONDS)
  public LinkedList<Integer> linkedList(
          LinkedListExecutionPlan executionPlan) {
    LinkedList<Integer> list = executionPlan.linkedList;

    list.remove(list.size() - 1);

    return list;
  }

TRY IT YOURSELF: You can find this source code here.

Now, let’s see the results:

That is pretty interesting, LinkedList and ArrayList have pretty much a equal behavior. Let’s see why they behave like that.

NOTE: Careful, if you run those tests multiple times, the result mightn’t be equal

ArrayList

Well, let’s use this example:

Now, we want to remove the index 6 of the list. So, ArrayList will do the following:

1. Set NULL the 6 position:

2. Change the array size variable from 7 to 6:

TIP: In a logic point of view, the ArrayList size is now 6, however, its pure array is 7. That means, the next time we need to add a new element at the end, is going to be fast, there is a position available.

We can conclude:

• In the worst case, we just need to set NULL the last position
• The amount of operations is 1.
• The complexity of this operation is O(1).

LinkedList

Well, let’s use this example:

Now, we want to remove the 6 index from the list. This is the process:

1.Move the Tail one index before:

2. Unlink the previous Tail from the new Tail:

3. The garbage collector is going to reclaim the unlinked node:

We can conclude:

• We only need to unlink the node from the Tail.
• The amount of operations is 1.
• The complexity of this operation is O(1).

Deleting from the Middle – remove(size / 2)

Now, let’s see the code for measuring how ArrayList and LinkedList behave when we need to remove from the middle:

  @Test
  public void remove_middle()
          throws RunnerException, IOException {
    Utils.populateToFile();

    Options opt = new OptionsBuilder()
            .include(
                    ArrayListVsLinkedListDeletionMiddleTest.class
                            .getSimpleName())
            .detectJvmArgs()
            .build();

    Collection<Integer> runResults = new Runner(opt)
            .run();

    Result arrayListResult = Utils
            .find("arrayList",
                    runResults);
    Result linkedListResult = Utils
            .find("linkedList",
                    runResults);

    assertThat(linkedListResult.getScore())
            .isLessThan(arrayListResult.getScore());
  }

  @Benchmark
  @Measurement(iterations = Utils.amountIterations, batchSize = 1, time = 1)
  @BenchmarkMode(Mode.SingleShotTime)
  @Fork(1)
  @Warmup(iterations = 10)
  @OutputTimeUnit(TimeUnit.MILLISECONDS)
  public ArrayList<Integer> arrayList(
          ArrayListExecutionPlan executionPlan) {
    ArrayList<Integer> list = executionPlan.arrayList;

    list.remove(list.size() / 2);

    return list;
  }

  @Benchmark
  @Measurement(iterations = Utils.amountIterations, batchSize = 1, time = 1)
  @BenchmarkMode(Mode.SingleShotTime)
  @Fork(1)
  @Warmup(iterations = 10)
  @OutputTimeUnit(TimeUnit.MILLISECONDS)
  public LinkedList<Integer> linkedList(
          LinkedListExecutionPlan executionPlan) {
    LinkedList<Integer> list = executionPlan.linkedList;

    list.remove(list.size() / 2);

    return list;
  }

TRY IT YOURSELF: You can find this source code here.

Now, let’s see the results:

Again, LinkedList and ArrayList have pretty much a equal behavior. Let’s see why they behave like that.

ArrayList

Well, let’s use this example:

Now, we want to remove the 3 index from the list. So, ArrayList will do the following:

1. Set NULL the 3 index:

2. Move the whole elements after 3 index one position before:

We can conclude:

• In the worst case, we need to move the half of the list one position before.
• The amount of operations is n/2.
• The complexity of this operation is O(n).

LinkedList

Well, let’s use this example:

Now, we want to remove the 3 index from the list. This is the process:

1. Move to the index from the closest reference (Head or Tail), in this case, Head is closest:

2. Removing the reference from the 3 index, to previous (2) and next (4):

3. Linking the 2 index to 4 index:

4. The garbage collector is going to reclaim the unlinked node:

We can conclude:

• In the worst case, we need to move from the head/tail to the index
• The amount of operations is n/2.
• The complexity of this operation is O(n).

Removing from the Start – remove(0)

Now, let’s see the code for measuring how ArrayList and LinkedList behave when we need to remove from the start:

  @Test
  public void remove_start()
          throws RunnerException, IOException {
    Utils.populateToFile();

    Options opt = new OptionsBuilder()
            .include(
                    ArrayListVsLinkedListDeletionStartTest.class
                            .getSimpleName())
            .detectJvmArgs()
            .build();

    Collection<RunResult> runResults = new Runner(opt)
            .run();

    Result arrayListResult = Utils
            .find("arrayList",
                    runResults);
    Result linkedListResult = Utils
            .find("linkedList",
                    runResults);

    assertThat(linkedListResult.getScore())
            .isLessThan(arrayListResult.getScore());
  }

  @Benchmark
  @Measurement(iterations = Utils.amountIterations, batchSize = 1, time = 1)
  @BenchmarkMode(Mode.SingleShotTime)
  @Fork(1)
  @Warmup(iterations = 10)
  @OutputTimeUnit(TimeUnit.MILLISECONDS)
  public ArrayList<Integer> arrayList(
          ArrayListExecutionPlan executionPlan) {
    ArrayList<Integer> list = executionPlan.arrayList;

    list.remove(0);

    return list;
  }

  @Benchmark
  @Measurement(iterations = Utils.amountIterations, batchSize = 1, time = 1)
  @BenchmarkMode(Mode.SingleShotTime)
  @Fork(1)
  @Warmup(iterations = 10)
  @OutputTimeUnit(TimeUnit.MILLISECONDS)
  public LinkedList<Integer> linkedList(
          LinkedListExecutionPlan executionPlan) {
    LinkedList<Integer> list = executionPlan.linkedList;

    list.remove(0);

    return list;
  }

TRY IT YOURSELF: You can find this source code here.

Now, let’s see the results:

LinkedList is faster, pretty fast. Let’s see why they behave like that.

ArrayList

Well, let’s use this example:

Now, we want to remove the 0 index from the list. So, ArrayList will do the following:

1. Set NULL the 0 index:

2. Move the whole elements one position before:

2. Change the array size variable from 7 to 6:

We can conclude:

• In the worst case, we need to move the whole elements minus one, one position before.
• The amount of operations is n-1.
• The complexity of this operation is O(n).

LinkedList

Well, let’s use this example:

Now, we want to remove the 0 index from the list. This is the process:

1. Move the Head one position forward:

2. Unlink the 0 index from the Head:

3. The garbage collector is going to reclaim the unlinked node:

We can conclude:

• We only need to unlink the node from the Head.
• The amount of operations is 1.
• The complexity of this operation is O(1).

Summary

Okey, this is a summary of the amount of operations in the deletion operations between ArrayList and LinkedList:

		ArrayList		LinkedList
Operation	Method	Best	Worst	Best	Worst
Remove From End	remove(size-1)	1	1	1	1
Remove From Middle	remove(size/2)	n/2	n/2	n/2	n/2
Remove From Start	remove(0)	n-1	n-1	1	1

Then, this is a summary of the Big O complexity in the deletion operations between ArrayList and LinkedList:

		ArrayList		LinkedList
Operation	Method	Best	Worst	Best	Worst
Remove From End	remove(size-1)	O(1)	O(1)	O(1)	O(1)
Remove From Middle	remove(size/2)	O(n)	O(n)	O(n)	O(n)
Remove From Start	remove(0)	O(n)	O(n)	O(1)	O(1)

Final Thought

As we saw, LinkedList is better than ArrayList when we talk about deletion operations.

However, ArrayList behaves similar in 2 of 3 deletion operations.

In the following post, we are going to talk about sort, get and iteration operations.

If you liked this post and are interested in hearing more about my journey as a Software Engineer, you can follow me on Twitter and travel together.