Class #17 – Recursion and More Sorting



1def fact(n):
2    if n == 1:
3        return 1
4    else:
5        return n * fact(n-1)


Test this code. What’s the answer? Why is that the answer? Trace through this code.

There is nothing really special happening here; we’re following the same rules we always have.


Don’t Panic

  • Before talking about the next few things, I want you to be very well aware that these algorithms are outside the scope of this course

  • I am showing you clever algorithms

  • Probably the most sophisticated we’ve seen so far

  • For now, don’t get distracted by all the details, just think about the big picture.

  • The takeaway from this is to get a feel for how neat algorithms can get.


High Level

  • An empty list, or a list with one thing in it, is already sorted, right?

  • If I take a sorted list of things less than a number x, and a sorted list of things greater than a number x, then I can just do this:

    [less than x] + [x] + [greater than x]

  • This will also result in a sorted list, right?

Less high level

  • You give me a list called in_list

  • If the list has only one element (or none), just return the list without doing anything
    • because it is a sorted list — DONE!

  • I pick an element from the list, which I’ll call the pivot

  • I partition the list by shuffling the elements around so that:
    • all elements less than pivot are to the left of pivot

    • all elements greater than pivot are to the right of it

  • I recursively call quicksort on two lists:
    • The list of stuff to the left of pivot

    • The list of stuff to the right of pivot

 1def partition(list, l, e, g):
 2    while list != []:
 3        head = list.pop(0)
 4        if head < e[0]:
 5            l = [head] + l
 6        elif head > e[0]:
 7            g = [head] + g
 8        else:
 9            e = [head] + e
10    return (l, e, g)
12def quicksort(list):
13    print(list)
14    if list == []:
15        return []
16    else:
17        pivot = list[0]
18        less, equal, greater = partition(list[1:],[],[pivot],[])
19        return quicksort(less) + equal + quicksort(greater)


High Level

  • An empty list, or a list with one thing in it, is already sorted, right?

  • We can merge two sorted lists together into a new, bigger sorter list.

Less high level

  • You give me a list, in_list with n elements

  • I divide in_list into n sublists, each having 1 element.

  • The sublists are now already sorted! (Any list with only one element is automatically sorted)

  • I now begin merging sublists, to produce bigger (but still sorted) sublists:
    • I compare the first element in one sublist to the first in the other

    • Whichever is smaller goes into the merged list

    • I now compare the second element in that list to the first element in the other list

    • … and so on…

  • I keep going until there is nothing left to merge (there’s only one big, sorted, list)

 1def merge(left, right):
 2    merged = []
 3    l=0
 4    r=0
 5    while l < len(left) and r < len(right):
 6        if left[l] <= right[r]:
 7            merged.append(left[l])
 8            l = l + 1
 9        else:
10            merged.append(right[r])
11            r = r + 1
13    if l < len(left):
14        merged.extend(left[l:])
15    else:
16         merged.extend(right[r:])
17    return merged
19def mergesort(list):
20    if len(list) <= 1:
21        return list
22    else:
23        midpoint = int( len(list) / 2 )
24        left = mergesort(list[:midpoint])
25        right = mergesort(list[midpoint:])
26        return merge(left,right)

Self Reference

  • Let’s say I have the following code:

    1a = [5, 6, 7]
  • What is the length of the list a?

  • What is in the list at index 3?

  • What’s at a[3][3]?

  • What’s at a[3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3][3]?

  • Remember how things got kinda’ interesting when we start dealing with self-references?

  • Is the following statement True of False?

    This statement is False

  • What happens when Pinocchio says

    My nose is going to grow now

Grelling-Nelson Paradox

  • Autological and Heterological words
    • An autological word is one that describes itself
      • English

      • unhyphenated

      • pentasyllabic

    • A heterological word is one that does not describe itself
      • French

      • hyphenated

      • monosyllabic

    • If you’re lost, just ask yourself: Is the word English English? Is the word unhyphenated unhyphenated? Is the word hyphenated hyphenated?

  • Seems reasonable to say that we can clearly categorize all the words as either autological or heterological.

  • Let’s put them into Set_A or Set_H

  • Where do we put the word round?

    Is the word round round?

  • Where do we put the word writable?

    Is the word writable writable?

  • Where do we put the word harmless?

    Is the word harmless harmless?

  • Where do we put the word autological?

    Is the word autological autological?

    • Hmmmmmm

    • Weird one…

    • A little different than the previous

    • If it is, it is.
      • If it’s autological, then it is autological and belongs in the set of autological words.

    • If it’s not, it’s not…
      • If it’s not autological, then it is heteroligical and it belongs in the set of heterological words.

  • Where do we put the word heterological?

    Is the word heteroligical heteroligical?

    • Let’s start by saying the answer to this question is yes.
      • If it is, then the word is describing itself.

      • Therefore, the word heterological must be heterological

      • What does it mean for a word to be heterological?

      • Well, for a word to be heterological, it must not be describing itself.

      • So if the word heteroligical is heterological, it must not be describing itself, and therefore it isn’t autological…

      • For the heterological to be autological, it must be heteroligical, which would make it not autological.

    • Is it then heteroligical?
      • So if it does not describe itself, then that means that the word does not describe itself.

      • Wait a min… that’s the definition of the word heterological…

      • If the word heterological is heteroligical, it is therefore autological, which would mean that it’s not heteroligical…

Halting Problem


Can you write a Python program that will look at a Python function and determine if the function might go into an infinite loop (does not halt)?

  • Let’s assume the answer to this activity is “yes” and see where that gets us.

1def halt(f, i):
2    if f(i) halts:
3        return True
4    else:
5        return False
  • Given this, I can easily do this:

1def problem0(f):
2    if halt(f, f) returns True:
3        # Go into an infinite loop
4        while True:
5            pass
6    else:
7        return False
  • This may seem weird (it is), but it’s totally legal, right?

  • What happens if I call problem0(problem0)?

  • Let’s see what happens:

    problem0(problem0) –> halt(problem0, problem0) –> problem0(problem0)

  • Ok, let’s simplify this by asking 2 small questions:
    • What would cause problem0(problem0) to go into an infinite loop?
      • That would mean that halt(problem0, problem0) returned True, which means halt said that the program halted.

      • Looking into halt, this happens when f(i) halts.

      • In our case, f is problem0 and i is problem0

      • So, when problem0(problem0) halts.

    • Wait… problem0(problem0) goes into an infinite loop when problem0(problem0) does not go into an infinite loop.

    • Ok, we’ve got a problem here. So let’s just ignore this and ask the other question and hope my day isn’t ruined.

    • What would cause problem0(problem0) to not go into an infinite loop?
      • That would mean that halt(problem0, problem0) returned False

      • This will happen when problem0(problem0) does not go into an infinite loop

    • I give up.


Why is this interesting?

  • Humans want to be able to ask the question

    Given these premises, can we get a specific conclusion?

  • This shows us that, given our computational model, there are some things that we just can not compute.