When to use a linked list over an array/array list

问题:

I use a lot of lists and arrays but I have yet to come across a scenario in which the array list couldn\'t be used just as easily as, if not easier than, the linked list. I was hoping someone could give me some examples of when the linked list is notably better.

回答1:

Linked lists are preferable over arrays when:

a) you need constant-time insertions/deletions from the list (such as in real-time computing where time predictability is absolutely critical)

b) you don\'t know how many items will be in the list. With arrays, you may need to re-declare and copy memory if the array grows too big

c) you don\'t need random access to any elements

d) you want to be able to insert items in the middle of the list (such as a priority queue)

Arrays are preferable when:

a) you need indexed/random access to elements

b) you know the number of elements in the array ahead of time so that you can allocate the correct amount of memory for the array

c) you need speed when iterating through all the elements in sequence. You can use pointer math on the array to access each element, whereas you need to lookup the node based on the pointer for each element in linked list, which may result in page faults which may result in performance hits.

d) memory is a concern. Filled arrays take up less memory than linked lists. Each element in the array is just the data. Each linked list node requires the data as well as one (or more) pointers to the other elements in the linked list.

Array Lists (like those in .Net) give you the benefits of arrays, but dynamically allocate resources for you so that you don\'t need to worry too much about list size and you can delete items at any index without any effort or re-shuffling elements around. Performance-wise, arraylists are slower than raw arrays.

回答2:

Arrays have O(1) random access, but are really expensive to add stuff onto or remove stuff from.

Linked lists are really cheap to add or remove items anywhere and to iterate, but random access is O(n).

回答3:

To add to the other answers, most array list implementations reserve extra capacity at the end of the list so that new elements can be added to the end of the list in O(1) time. When the capacity of an array list is exceeded, a new, larger array is allocated internally, and all the old elements are copied over. Usually, the new array is double the size of the old one. This means that on average, adding new elements to the end of an array list is an O(1) operation in these implementations. So even if you don\'t know the number of elements in advance, an array list may still be faster than a linked list for adding elements, as long as you are adding them at the end. Obviously, inserting new elements at arbitrary locations in an array list is still an O(n) operation.

Accessing elements in an array list is also faster than a linked list, even if the accesses are sequential. This is because array elements are stored in contiguous memory and can be cached easily. Linked list nodes can potentially be scattered over many different pages.

I would recommend only using a linked list if you know that you\'re going to be inserting or deleting items at arbitrary locations. Array lists will be faster for pretty much everything else.

回答4:

Algorithm           ArrayList   LinkedList
seek front            O(1)         O(1)
seek back             O(1)         O(1)
seek to index         O(1)         O(N)
insert at front       O(N)         O(1)
insert at back        O(1)         O(1)
insert after an item  O(N)         O(1)

ArrayLists are good for write-once-read-many or appenders, but bad at add/remove from the front or middle.

回答5:

The advantage of lists appears if you need to insert items in the middle and don\'t want to start resizing the array and shifting things around.

You\'re correct in that this is typically not the case. I\'ve had a few very specific cases like that, but not too many.

回答6:

Those are the most common used implementations of Collection.

ArrayList:

• insert/delete at the end generally O(1) worst case O(n)

• insert/delete in the middle O(n)

• retrieve any position O(1)

LinkedList:

• insert/delete in any position O(1) (note if you have a reference to the element)

• retrieve in the middle O(n)

• retrieve first or last element O(1)

Vector: don\'t use it. It is an old implementation similar to ArrayList but with all methods synchronized. It is not the correct approach for a shared list in a multithreading environment.

HashMap

insert/delete/retrieve by key in O(1)

TreeSet insert/delete/contains in O(log N)

HashSet insert/remove/contains/size in O(1)

回答7:

It all depends what type of operation you are doing while iterating , all data structures have trade off between time and memory and depending on our needs we should choose the right DS. So there are some cases where LinkedList are faster then array and vice versa . Consider the three basic operation on data structures.

• Searching

Since array is index based data structure searching array.get(index) will take O(1) time while linkedlist is not index DS so you will need to traverse up to index , where index <=n , n is size of linked list , so array is faster the linked list when have random access of elements.

Q.So what\'s the beauty behind this ?

As Arrays are contiguous memory blocks, large chunks of them will be loaded into the cache upon first access this makes it comparatively quick to access remaining elements of the array,as much as we access the elements in array locality of reference also increases thus less catch misses, Cache locality refers to the operations being in the cache and thus execute much faster as compared to in memory,basically In array we maximize the chances of sequential element access being in the cache. While Linked lists aren\'t necessarily in contiguous blocks of memory, there\'s no guarantee that items which appear sequentially in the list are actually arranged near each-other in memory, this means fewer cache hits e.g. more cache misses because we need to read from memory for every access of linked list element which increases the time it takes to access them and degraded performance so if we are doing more random access operation aka searching , array will be fast as explained below.

• Insertion

This is easy and fast in LinkedList as insertion is O(1) operation in LinkedList (in Java) as compared to array, consider the case when array is full, we need to copy contents to new array if array gets full which makes inserting an element into ArrayList of O(n) in worst case, while ArrayList also needs to update its index if you insert something anywhere except at the end of array , in case of linked list we needn\'t to be resize it, you just need to update pointers.

• Deletion

It works like insertions and better in LinkedList than array.

回答8:

Hmm, Arraylist can be used in cases like follows I guess:

1. you are not sure how many elements will be present
2. but you need to access all the elements randomly through indexing

For eg, you need to import and access all elements in a contact list (the size of which is unknown to you)

回答9:

Use linked list for Radix Sort over arrays and for polynomial operations.

回答10:

1) As explained above the insert and remove operations give good performance (O(1)) in LinkedList compared to ArrayList(O(n)). Hence if there is a requirement of frequent addition and deletion in application then LinkedList is a best choice.

2) Search (get method) operations are fast in Arraylist (O(1)) but not in LinkedList (O(n)) so If there are less add and remove operations and more search operations requirement, ArrayList would be your best bet.

回答11:

I think that main difference is whether you frequently need to insert or remove stuff from the top of the list.

With an array, if you remove something from the top of list than the complexity is o(n) because all of the indices of the array elements will have to shift.

With a linked list, it is o(1) because you need only create the node, reassign the head and assign the reference to next as the previous head.

When frequently inserting or removing at the end of the list, arrays are preferable because the complexity will be o(1), no reindexing is required, but for a linked list it will be o(n) because you need to go from the head to the last node.

I think that searching in both linked list and arrays will be o(log n) because you will be probably be using a binary search.

回答12:

I did some benchmarking, and found that the list class is actually faster than LinkedList for random inserting:

using System;
using System.Collections.Generic;
using System.Diagnostics;

namespace ConsoleApplication1
{
    class Program
    {
        static void Main(string[] args)
        {
            int count = 20000;
            Random rand = new Random(12345);

            Stopwatch watch = Stopwatch.StartNew();
            LinkedList<int> ll = new LinkedList<int>();
            ll.AddLast(0);
            for (int i = 1; i < count; i++)
            {
                ll.AddBefore(ll.Find(rand.Next(i)),i);

            }
            Console.WriteLine(\"LinkedList/Random Add: {0}ms\", watch.ElapsedMilliseconds);

            watch = Stopwatch.StartNew();
            List<int> list = new List<int>();
            list.Add(0);
            for (int i = 1; i < count; i++)
            {
                list.Insert(list.IndexOf(rand.Next(i)), i);

            }
            Console.WriteLine(\"List/Random Add: {0}ms\", watch.ElapsedMilliseconds);

            Console.ReadLine();
        }
    }
}

It takes 900 ms for the linked list and 100ms for the list class.

It creates lists of subsequent integer numbers. Each new integer is inserted after a random number which is already in the list. Maybe the List class uses something better than just an array.

回答13:

In reality memory locality has a huge performance influence in real processing.

The increased use of disk streaming in \"big data\" processing vs random access shows how structuring your application around this can dramatically improve performance on a larger scale.

If there is any way to access an array sequentially that is by far the best performing. Designing with this as a goal should be at least considered if performance is important.