gistfile1.txt

possible assignments:

- add tests to check the behaviour is correct.
  don't compare it to python's output, just check that each element is greater than the
  next, with the compare function

- add profiling code, and run it with much, much larger lists ~5000 elements
  which one is fast, slow?

- quick sort picks a random pivot, is this a good choice? what happens on larger lists
  does the time vary? what happens if you take a bunch of points (top, middle, bottom)
  and use an average of them to split the array into two lists

- quicksort can be done in-place, without creating extra lists. can you write a new 
  quicksort that doesn't create temporary arrays:

  hint: your quicksort will take three arguments, the array, and the range to sort within
  i.e qsort_inplace(array, lo, hi) where you only sort x where lo <= x < hi, and do
  qsort_inplace(array, 0, len(array)) to begin with  

- make the merge sort smarter when splitting lists, this is called "adaptive merge sort"
  i.e split the list into ascending and decending sequences and merge them

- fill up the hash table with entries (it only has 8 buckets), what happens, do the entries
  split evenly amongst the buckets, or does one bucket get really big. what happens
  to the performance