Hash＃has_key？和Array＃index的性能

Question

I frequently check if a specific value is in a large array. I can do this by Array#index . To make this more efficient, I created a hash of the array values and called Hash#has_key? :

• Method 1

   arr = ["a","b","c","d"] arr.index("c") 

• Method 2

   h = {"a"=> true, "b"=> true, "c"=> true, "d"=> true} h.has_key?("c") 

But I noticed ruby throws an exception if a key isn't in a given hash. I'm wondering what the relative performance of the two methods is.

Answer 1

To answer your question, "Method 2" should be faster. Now, that is a very loaded statement which partially depends on the very nature of hashes (eg collisions when inserting ).

However, for your specific use case, I think arrays and hashes are both the "wrong tool for the job". In general, if you're using a hash to check unique set existence (hint hint), use a set .

One final thought, which may or may not be valuable depending on how contrived your example is. If you're storing some finite set of ordered values ('a'-'d' in your example) an array is definitely the way to go. Why? Because you can easily map the values of your alphabet to an array index (eg a maps to 0, b maps to 1 and so forth) by, in your case, converting the letters to ascii and subtracting to get their desired location. This would give you an O(1) lookup time.

Answer 2

Ruby has a construct in the standard library that gives you what you want: O(1) lookups using #include? .

Set class documentation

require 'set'
arr = ["a","b","c","d"]
set = Set.new(arr)
set.include?("c")

Note however that this only works if you don't care about duplicate elements (but I am assuming that's the case based on your 2nd method, which also depends on that assumption).