Sorting and searching at the library

The algorithms librarians use for sorting and searching physical books are a little different from what you learned in school. The reasons for these differences are surprising.

Transcript

1. by jason orendorff for hack day, 11 august 2012 at the library Saturday, August 11, 12
2. searching Saturday, August 11, 12
3. books = [ "001.4 Graphs", "001.56 Symbols", ..., "998.9 Antarctica" ] Saturday, August 11, 12
4. "664.75 Cookies" Saturday, August 11, 12
5. books.index("664.75 Cookies") Saturday, August 11, 12
6. binary search is optimal Saturday, August 11, 12
7. a sorted array Saturday, August 11, 12
8. Saturday, August 11, 12
9. binary search in practice Saturday, August 11, 12
10. binary search in practice Saturday, August 11, 12
11. Saturday, August 11, 12
12. • humans are not machines? Saturday, August 11, 12
13. • humans are not machines Saturday, August 11, 12
14. • humans are not machines • this is not the binary search problem? Saturday, August 11, 12
15. • humans are not machines • this is not the binary search problem Saturday, August 11, 12
16. • humans are not machines • this is not the binary search problem • the cost model is wrong Saturday, August 11, 12
17. • humans are not machines • this is not the binary search problem • the cost model is wrong • also: the library cheats Saturday, August 11, 12
18. a little extra Saturday, August 11, 12
19. Saturday, August 11, 12
20. sorting Saturday, August 11, 12
21. how to sort cart.sort! cart.sort() (sort cart) Array.Sort(cart); Collections.Sort(cart); std::sort(cart.begin(), cart.end()); Saturday, August 11, 12
22. ).sort() ( Saturday, August 11, 12
23. ).sort() ( TypeError Saturday, August 11, 12
24. merge sort is optimal Saturday, August 11, 12
25. merge sort is optimal *up to a constant factor * Saturday, August 11, 12
26. Saturday, August 11, 12
27. Saturday, August 11, 12
28. merge sort Saturday, August 11, 12
29. merge sort • divide (in half) Saturday, August 11, 12
30. merge sort • divide (in half) • sort halves Saturday, August 11, 12
31. merge sort • divide (in half) • sort halves • merge Saturday, August 11, 12
32. merge sort • divide (in half) • sort halves • merge bucket sort Saturday, August 11, 12
33. merge sort • divide (in half) • sort halves • merge bucket sort • divide (in bins) Saturday, August 11, 12
34. merge sort • divide (in half) • sort halves • merge bucket sort • divide (in bins) • sort bins Saturday, August 11, 12
35. merge sort • divide (in half) • sort halves • merge bucket sort • divide (in bins) • sort bins • merge done! Saturday, August 11, 12
36. Insertion sort (it’s not optimal, just the fastest) Saturday, August 11, 12
37. Saturday, August 11, 12
38. Saturday, August 11, 12
39. Saturday, August 11, 12
40. Saturday, August 11, 12
41. Saturday, August 11, 12
42. Saturday, August 11, 12
43. library.nashville.org/volunteer Saturday, August 11, 12
44. Saturday, August 11, 12