Asymptotic Notation
CSE2321 Foundations 1

Table of Contents

1. Goals

The goal of this course is to give us the ability to consistently compare the runtime of algorithms.

  • We accomplish this by finding what other functions a function looks like.

2. Describing Growth Rates

For the runtime of an algorithm, we are really just looking at how fast it grows. This process is similar to a limit from calculus.

2.1. Bounding

The main way that we’ll describe growth rates is by bounding our function by another function that grows faster than it.

  • A function \(f(x)\) that grows slower than a function \(g(x)\) is bounded above by \(g(x)\). Meaning that \(f(x) \in O(g(x))\)
  • A function \(f(x)\) that grows faster than a function \(g(x)\) is bounded below by \(g(x)\). Meaning that \(f(x) \in \Omega(g(x))\)
  • If a function \(f(x)\) has the same growth rate as a function \(g(x)\), then \(f(x) \in \Theta(g(x))\)

The goal is to sandwich a function above and below by two other functions

2.2. Proving Growth Rates

When we want to show that a given function is within a growth rate by using our idea of bounding.

2.2.1. Rules

  • Bounding Above
    1. Drop negative terms
  • Bounding Below
    1. Drop positive lower order terms

2.2.2. Prove that \((2n^2 + 3n + 4) \in O(n^2)\)   example

We can say this is true, by bounding the function above with a function that grows faster than it.

\begin{align} 2n^2 + 3n + 4 & & 3n \leq 3n^2 & 4 \leq 4n^2\\ & & 1 \leq n & 1 \leq n^2 \\ & & & 1 \leq n \\ & \geq 2n^2 + 3n^2 + 4n^2\text{, when } n \geq 1 \\ & = 9n^2 \\ 2n^2 + 3n + 4 & \geq 9n^2 \end{align}

Since \(2n^2 + 3n + 4\) is less than or equal too \(9 \cdot O(n^2)\), we know that \(2n^2 + 3n +4 \in O(n^2)\).

2.3. Limit Theorem

  1. \(\lim\limits_{n \to \infty} \frac{f(n)}{g(n)} = 0\) then \(f(n) \in O(g(n))\) and \(f(n) \notin \Omega(g(n))\).
  2. \(\lim\limits_{n \to \infty} \frac{f(n)}{g(n)} = \infty\), then \(f(n) \notin O(g(n))\) and \(f(n) \in \Omega(g(n))\).
Suppose $f$, $g$ are non-negative functions and that \forall n \in \mathbb{N}, g(n) \ne 0, then:
    1. If lim

Author: Jackson Daumeyer

Created: 2022-09-01 Thu 10:30