Group Anagrams

Write `group_anagrams(words)` that groups the input into lists of mutual anagrams. Return a list of groups; order within a group and between groups does not matter. group_anagrams(["eat", "tea", "tan", "ate", "nat", "bat"]) -> [["eat", "tea", "ate"], ["tan", "nat"], ["bat"]] Every solution here hashes each word to a key that anagrams share. Which key you pick is the interview: sorting each word costs O(k log k) per word, a character-count tuple costs O(k). Say which you chose and why.