--- title: Eboard 35 An introduction to sorting number: 35 section: eboards held: 2019-04-29 link: true --- CSC 151 2019S, Class 35: An introduction to sorting ==================================================== _Overview_ * Preliminaries * Notes and news * Upcoming work * Extra credit * Questions * The problem of sorting * Writing sorting algorithms * Some examples * Formalizing the problem Preliminaries ------------- ### News / Etc. * I brought food-like substances. * Mentor sessions Thursday 7-8 p.m., Thursday 8-9 p.m. * Today is a "talk day" rather than a lab day. * Congrats to your classmate who tied for second in the Kinetic Sculpture Competition! * Welcome to any prospective students. Good luck on your decisions! ### Upcoming work * [Project](../project) due Tuesday! * [Rubric](../project/rubric.txt) available for your perusal. * Send me code, instructions for running, and report. * The parts of the report are specified in the rubric. I'd expect an introdution that describes the problem, a description of the key algorithms you wrote, a description of what you did with the results of those algorithms, and some conclusions about what you might have learned. Four sections, 1-2 pargraphs each. * Reading for Wednesday: [Algorithms for sorting lists and vectors](../readings/sorting) * No lab writeup * Friday's quiz: Searching and sorting * Flash cards Wednesday: Searching and sorting * Our last flash cards! * [Exam 3](../exams/exam03) due next Tuesday. * It is not "officially" released until this Wednesday; I still have time to correct typos and infelicities. * Prologue due Friday night. ### Extra Credit _I would certainly appreciate suggestions of other extra credit activities (preferably via email)._ #### Extra credit (Academic/Artistic) * **Tomorrow**: CS Table, Tuesday, Facebook Data, Noon, Middle PDR * Three talks by Prof. Dr. Yvonne Foerster () * Wednesday: May 1, 4:30-6pm, HSSC S3325: _Beyond the Anthropocene: Technology, Innovation, and the (Post-)Human Condition_ * Thursday, May 2, Noon-12:50pm, HSSC N3110 _Degrees of Freedom: Embodiment, Neuroplasticity, and the Need for a Critical Neuroscience_ (Lunch and beverages provided) * Friday, May 3, Noon-12:50pm, Bucksbaum 152: _Designing Future Bodies: Fashion and Technology_ (Lunch and beverages provided) #### Extra credit (Peer) * **New**: Sunday, 2pm, Herrick: Singers vs. Orchestra #### Extra credit (Wellness) #### Extra credit (Wellness, Regular) * **Today**: 30 Minutes of Mindfulness at SHAW every Monday 4:15-4:45 * Any organized exercise. * 60 minutes of some solitary self-care activities that are unrelated to academics or work. Your email reflection must explain how the activity contributed to your wellness. * 60 minutes of some shared self-care activity with friends. Your email reflection must explain how the activity contributed to your wellness. #### Extra credit (Misc) ### Other good things * The Grinnellian: Student concert, next Saturday. ### Questions The problem of sorting ---------------------- * We like to sort; it is useful to have things in order. * It makes binary search easier. * It makes our remove-duplicates algorithm much faster (or at least easier). * A really inefficient way to do max or min. (Nicer if you need like the first three or last three or ....) * Can help with hash tables. * Writing sorting algorithms is therefore an important task for computer scientists. * That will be our task for the next few days. Writing sorting algorithms -------------------------- * Look at similar algorithms we've written in the past and adapt. * Do things by hand and then try to generalize. Some examples ------------- Insertion sort * Split our world into sorted and unsorted. * Initially, the first element is sorted and everything else is unsorted. * Repeatedly take the first element of unsorted and put it in the correct place in sorted. Radix * Put all the A's together * Put all the B's together * Put all the C's together * Put all the D's together * In each group, do the same thing with the second letter * And so on and so forth * Can improve with knowledge of the distribution Radix plus insertion * Divide into three groups, smaller, middle, larger. Then run insertion. Selection * Conceptually we have two lists: sorted and unsorted * Repeatedly find the smallest thing in unsorted and add it to the end of sorted. Timsort will come on Friday Clownsort / Permutation sort * Randomly permute the list * See if it's sorted * If so, you're done * If not, start over Parallel bubble sort * Compare to the person next to you (people in odd positions to people in even positions) and swap if out of order. Machine learning sort * Give the problem to a machine learning algorithm * Use what it comes up with Analysis -------- Insertion sort * Putting the first thing in the new list costs us about 1. * Putting the second thing the new list costs us about 2 (in the worst case) * Putting the third thing the new list costs us about 3 (in the worst case) * Worst case 1+2+3+4+...+n. = n*(n+1)/2. An "n squared" algorithm * Idea: Instead of comparing to neighboring elements, use binary search to find the location. * For binary search to work, things must be sorted. * For binary search to work, they must be in an array. * Once we find where it goes, we have to shift things over. Note: With searching, we found that it was useful to think about limiting the form of the values (sortted) and using a "divide and conquer" algorithm. Improving sorting algorithms ---------------------------- We've seen that divide-and-conquer was a succesful strategy for searching; can we use it for sorting? * Divide the list in half (how?) * Do something with each half (what?) * Put them back together (how?) Mergesort * Divide the list into two halves, positionally * Sort each half (using recursion) * Merge the two together, by repeatedly grabbing the smaller of the first remaining element in each half. Median sort * Divide the list into those less than the median and greater than the median * Sort the two halves * Append together Quicksort * Median sort, using a randomly chosen median * Base case: 1 element Formalizing the problem -----------------------