Jared NielsenAfter Big O, the second most terrifying computer science topic might be recursion. Don’t let the memes scare you, recursion is just recursion. It’s very easy to understand and you don’t need to be a 10X developer to do so. In this tutorial, you’ll learn the fundamentals of calculating Big O recursive time complexity.
This article originally published on jarednielsen.com
What Problem(s) Does Recursion Solve?
- Recursion allows us to write functions that are compact and elegant.
What Problem(s) Does Recursion Create?
Recursion can easily exceed the maximum size of the call stack.
Recursion can make the program harder to understand not only for your collaborators, but for your future self
What is Recursion?
In computer science, recursion occurs when a function calls itself within its declaration.
For example:
const loop = () => loop();
If you run this in your browser console or using Node, you’ll get an error.
Why?
Too much recursion!
const loop() is just that, a constant loop.
🔁
We use recursion to solve a large problem by breaking it down into smaller instances of the same problem.
To do that, we need to tell our function what the smallest instance looks like.
If you recall, with proof by induction we need to establish two things:
- base
- induction
Recursion is similar. We also need to establish a base case but rather than induction, we establish the recursive case.
We use the recursive case to break the problem down into smaller instances.
We use the base case to return when there are no more problems to be solved.
For example. a family on vacation:
const fighting = patience => {
if (patience <= 0) {
return "If you don’t stop fighting, I will turn this car around!"
}
return fighting(patience - 1);
};
The kids are fighting in the backseat.
Dad is driving and quickly losing his patience.
Our recursive case is the constant fighting.
Our base case is dad’s patience when it runs out.
🚗
Recursion: Factorial
The classic example of recursion is computing the factorial of a given number.
What’s a factorial?
A factorial is the product of all positive integers less than or equal to n.
We write that as n!.
For example, 5!:
5 * 4 * 3 * 2 * 1 = 120
Here’s an iterative factorial in JavaScript:
const factorial = num => {
if (num === 0 || num === 1) {
return 1;
}
for (let i = num - 1; i >= 1; i--) {
num *= i;
}
return num;
};
And here it is refactored with recursion:
const factorial = num => {
if (num == 0 || num === 1) {
return 1;
} else {
return (num * factorial(num - 1));
}
};
Every call to factorial() again calls factorial(), but decreases the value of num by 1, until the base case is met and 1 is returned.
Fibonacci
Fibonacci is a sequence of numbers where each number is the sum of the preceding two.
It starts like this...
0 1 1 2 3 5 8 13 21 34 55 89 144…
Fibonacci algorithms are standard challenges for beginners and technical interviews. There are many ways to solve a Fibonacci algorithm and each reveals the depth of your knowledge.
Let’s dive in!
Iterative Fibonacci
Before we get to recursion, let’s look at an iterative solution to the problem.
Given an integer, n, calculate the sum of a Fibonacci sequence.
const fiberative = n => {
let arr = [0, 1];
for (let i = 2; i < n + 1; i++){
arr.push(arr[i - 2] + arr[i -1])
}
return arr[n];
};
What is the order of fiberative()?
O(n).
Why?
Our solution is iterative. We perform n operations.
If you want to go deeper, check out Big O Linear Time Complexity
Recursive Fibonacci
Now let’s implement our algorithm using recursion.
const fibonaive = n => {
if (n <= 0) {
return 0;
} else if (n === 1) {
return 1;
};
return fibonaive(n - 1) + fibonaive(n - 2);
};
What’s the order of fibonaive()?
Spoiler: it’s not good.
That’s why this approach is referred to as the naive implementation.
Let’s get informed.
Calculating Recursive Time Complexity
Let’s make a small adjustment to fibonaive() for the purpose of illustration:
const fibonot = n => {
if (n <= 0) {
return 0;
} else if (n === 1) {
return 1;
};
return fibonot(n - 1) + fibonot(n - 1);
};
☝️ We only modified the last line so that fibonot() is now balanced.
What’s happening in our function?
Every time we call fibonot(), we call fibonot() twice.
In each of those calls, we subtract 1 from n.
How many times does this happen?
n
We call fibonot() until the value of n is less than or equal to 0, or equal to 1, then we return without a call.
Every invocation of fibonot() creates two branches by calling itself twice.
Our branches are creating a tree.
With each iteration, the value of n becomes smaller until one of our base conditions is met.
So the depth of our tree is n.
Let’s map out the calls:
Do you see a pattern?
Where have we seen this, or something like it, before?
🤔
| Exponent | Power |
|---|---|
| 2^3 | 8 |
| 2^2 | 4 |
| 2^1 | 2 |
| 2^0 | 1 |
So what’s the order of fibonot()?
O(2^n)
Exponential!
Ew!
As a rule of thumb, when calculating recursive runtimes, use the following formula:
branches^depth
Where branches are the number of recursive calls made in the function definition and depth is the value passed to the first call.
In the illustration above, there are two branches with a depth of 4.
Let’s return to fibonaive().
What’s its Big O?
For our purposes, it’s O(2^n).
Technically, it’s O(1.6^n).
Why?
Let’s plant a tree! 🌲
What do you see?
Unlike fibonot() above, our tree is not balanced.
How many leaves are there on the tree?
We could count them by hand, but we’re problem solvers.
Math O’Clock 🧮🕒
Fibonacci is also expressed using the following formula:
F(n) = F(n -1) + F(n - 2)
Let’s use this formula to solve for x
x^n = x^(n -1) + x^(n - 2)
We first divide both sides by x^(n - 2)
x^2 = x + 1
Subtract 1 from both sides
x^2 - 1 = x
Subtract x from both sides
x^2 - 1 - x = 0
Where have we seen this, or something like it, before?
🤔
It’s a quadratic equation!
Quadratic equations follow the form:
ax^2 + bx + c = 0
We can use the quadratic formula to solve for x.
Let’s plug in our values:
- (-1) + sqrt((-1)^2 - 4 * 1 * (-1)) / 2 * 1
First, let’s simplify the numerator.
A negative negative is positive, so:
1 + sqrt((-1)^2 + 4 * 1 * 1)
A negative integer raised to a power is positive, so:
(-1)^2 = 1
Leaving us with the following:
1 + sqrt(1 + 4 * 1 * 1)
If we simplify the terms of the numerator:
1 + sqrt(5)
And simplify the terms of the denominator:
x = (1 + sqrt(5)) / 2
Which is equal to ~1.6.
AKA φ
AKA Phi
AKA Binet’s formula.
AKA the Golden Ratio.
☝️
“Hold up there, mister”, I hear you say.
“What about the other half of the quadratic formula?”
Good eye! 🕵️
You noticed that the quadratic formula results in two solutions, signified by the plus-minus sign.
Each solution charts the x-intercept of a parabola.
But we’re not interested in negative values, so we can stop with one solution.
Big O Recursive Space Complexity: The Final Frontier 🚀
If the time complexity of our recursive Fibonacci is O(2^n), what’s the space complexity?
We'll answer that question in the next tutorial.
Want to stay in the loop? I write a weekly newsletter about programming, problem solving and lifelong learning. Sign up for The Solution
