This question already has answers here: Big-O summary for Java Collections Framework implementations? [closed] (4 answers) Closed 8 years ago.

Question pretty much says it all. Specifically, I would like the Big-O of all the methods within a structure, aside from the usual. The docs say very little about this.

Addennum

For those who are voting to close, I am not interested in the basic add, remove, iterator, etc Those sources are fine for regularly used methods, but I am more interested in the algorithmic efficiency of the rest of the pile.

For example, what is the efficiency of开发者_StackOverflow中文版 TreeMap.keySet()?

---

Java Collections Algorithm Efficiencies: Source

ArrayList

• get, set: O(1)
• add, remove: O(n)
• contains, indexOf: O(n)

HashMap

• get, put, remove, containsKey: O(1)

HashSet

• add, remove, contains: O(1)

LinkedHashSet

• add, remove, contains: O(1)

LinkedList

• get, set, add, remove (from either end): O(1)
• get, set, add, remove (from index): O(n)
• contains, indexOf: O(n)

PriorityQueue

• peek: O(1)
• add, remove: O(logn)

TreeMap

• remove, get, put, containsKey: O(logn)

TreeSet

• add, remove, contains: O(logn)

---

I would review the standard Big-O information about algorithms in Introduction_to_Algorithms.

---

I am not sure is there a spreadsheet or other list, but you can get this information in docs for every structure, for example TreeSet:

This implementation provides guaranteed log(n) time cost for the basic operations (add, remove and contains).

---

There really cannot be. Each JVM can have it's own runtime, with its own implementations. A few methods have specified O() characteristics, but the rest are whatever Sun, or Oracle, or IBM, or Apple, or Azul, does.