CSCI H343 Data Structures Fall 2023

Lecture: Course Introduction

In the beginning was the Tao. The Tao gave birth to Space and Time. Therefore Space and Time are the Yin and Yang of programming.

Programmers that do not comprehend the Tao are always running out of time and space for their programs. Programmers that comprehend the Tao have enough time and space to accomplish their goals.

How could it be otherwise? – The Tao of Programming

Introduce myself

Introduce course AI’s

• Tianyu Chen

Course web page, syllabus, schedule (due dates, etc.)

https://iudatastructurescourse.github.io/course-web-page-fall-2023/

Participation:

• Write your name on a “name tent” and place it at the front of your desk

• Slack: ask and answer questions if not on slack, send me email

Suggested weekly schedule (4 credits * 3 = 12 hours of outside-class work)

• Before each lecture, read the assigned chapters. Do the textbook exercises as you read the chapters. The lecture should feel like a review with a few extra tidbits. Humans learn by repetition. If you don’t read, then you’re going to have a hard time understanding the lecture.

• Labs are every week. They will include a mixture of programming exercises, lecture review, help with homework and projects, and 15 minute quizzes.

• Homework (textbook exercises), 6 total for the semester

• Projects, about 1 per month, groups of 2-4 students. To complete the projects, you’ll typically need a firm understanding of the lectures from the previous two weeks and the Homework.

Overview of course content

• prerequisites:
  • C211 Intro. to CS basic programming skills, strings, lists, dicts.
  • C212 Software Systems: object-oriented programming, Java
  • C241 Discrete Math: the mathematics of programming
• postrequisites:
  • B403: Intro. to Algorithm Design and Analysis
  • B461: Database Concepts
  • B481: Interactive Graphics
  • P434: Distributed Systems
  • P436: Intro. to Operating Systems
  • P465: Software Engineering for Info. Systems I
• efficient procedures for solving problems on large data: U.S. highway system, Human genome, Social network/facebook
• tools of the trade, basic ways to organize information on computers
• real implementations in Java, testing on small to large inputs
• Many job interview questions are drawn from data structures & algorithms
• schedule on course web site, overview of projects (1 every few weeks)

Software Tools

• IntelliJ: Learn to use an Integrated Development Environment (Text Editor + Debugger)

• Autograder for submitting assignments

• Optional: learn github for managing group work

• ChatGPT/AI: you can use it but beware that if you use it to avoid learning, then you are likely to perform poorly on the midterm exam and final exam (60% of the course grade).

Lab 1: Testing Linear Search and Binary Search

Project 1: Flood It!

• Choose a team

Appetizer: The Maximum Subsequence Problem

Find the section of the Tecumseh Trail Marathon with the greatest net change in altitude. Show TecumsehMarathon.pdf.

change:   -25, -100, +100, -25,-75,-125,+100,+100,-25, +75,   -50, -100
			0     1     2    3   4    5    6    7   8    9   10    11

We want the contiguous subsequence with the largest sum. This is also known as the Maximum Subarray Problem.

• Brute force: try every pair of locations

  The time it takes, given an input array of size n, is

    n choose 2 = n * (n-1) / 2
  

  which is roughly n².

• Divide-and-conquer: Cut the array in half, then consider three cases:

  1. max-subarray is to the left of the cut
  2. max-subarray is to the right of the cut
  3. max-subarray crosses the cut

  Cases 1&2 are the same problem but smaller -> recursion Case 3 needs work:

  • any crossing subarray is made of a left and right part that meet in the middle
  • finding a max-subarray with one end held constant is easy: start at the constant end and keep growing the subarray toward the other end, keeping track of the max sum and index seen so far.

Analysis of the time complexity:

The crossing-subarray is linear, that is, c * m where m is the size of the subarray.

Recursion tree:

                c*n                     = c*n
            /         \
       c*n/2          c*n/2             = c*n
      /    \          /    \
    c*n/4  c*n/4   c*n/4   c*n/4        = c*n
    ...

What’s the height?

At the bottom, at a leaf, we have n/(2^h) = 1. We want to solve for h, so we isolate the 2^h term on one side.

n = 2^h

Then we take the log of both sides. (Recall that log (2^x) = x for any x.)

log n = h

Total work is c * n * lg n, so the time for this divide-and conquer algorithm for maximum-subarray is is roughly

n lg n