Experienced Computer Scientists analyze and solve computational problems at a level of abstraction that is beyond that of any particular programming language. This two-part course builds on the principles that you learned in our Principles of Computing course and is designed to train students in the mathematical concepts and process of "Algorithmic Thinking", allowing them to build simpler, more efficient solutions to real-world computational problems. In part 1 of this course, we will study the notion of algorithmic efficiency and consider its application to several problems from graph theory. As the central part of the course, students will implement several important graph algorithms in Python and then use these algorithms to analyze two large real-world data sets. The main focus of these tasks is to understand interaction between the algorithms and the structure of the data sets being analyzed by these algorithms. Recommended Background - Students should be comfortable writing intermediate size (300+ line) programs in Python and have a basic understanding of searching, sorting, and recursion. Students should also have a solid math background that includes algebra, pre-calculus and a familiarity with the math concepts covered in "Principles of Computing".
Illustrating "Big O"
Loading...
From the course by Rice University
Algorithmic Thinking (Part 1)
246 ratings
Explore Related Videos
At Coursera, you will find the best lectures in the world. Here are some of our personalized recommendations for you
From the lesson
Module 2 - Core Materials
Asymptotic analysis, "big O" notation, pseudocode, breadth-first search
Meet the Instructors
Luay NakhlehAssociate Professor
Computer Science; Biochemistry and Cell Biology
Scott RixnerProfessor
Computer Science
Joe WarrenProfessor
Computer Science
Okay, so now that we have these asymptotic notations Big O, omega and theta.
Now we can use these kind of, of notations and their definitions to put upper
bounds for example, on the running time of functions, and use these to give us a bet,
an id, to give us a good idea of how the algorithm is going to behave.
When we use an upper bound,
we are going to say that the algorithm is not going to get worse than this, or
the running time is not going to get worse than this.
When we put an upper on we put a lower bound,
we are basically saying, this is the best the algorithm can do.
And usually, the algorithms do not necessarily have even to
achieve these running times.
So if I put a lower bound of 2 to the N on an algorithm, it doesn't mean that
it's going to, it's going to, there are instances it's going to reach 2 to the n.
It might be that for
every instance of the input is going to be doing much more than 2 to the n.
But still, that makes 2 to the n,
it doesn't violate the fact that 2 to the n is a lower bound.
Now, of course, the, the, one of the issues that we have to make sure or to,
to make certain that everyone understands, is that when we put an upper bound or
lower bound, if all I'm saying is that give me a lower bound and
the running time of an algorithm, of course you know, one can say 0,
every algorithm is going to take more than 0 seconds, right?
Or more than, it's going to perform more than zero operations.
Now that is not an interesting lower bound.
Even though we're saying we want lower bounds and
upper bounds, we are often implicitly asking for the tightest possible, okay?
For the tightest running time possible.
And when we, let's now go back to the, to one of these which we are going to be
using most of the time in this course, which is the big connotation.
And just a, a, a reminder,
that we say that F of N is O of G of N,
if there exists two constants such
that F of N is smaller than C times G of N for
every n greater than n 0.
Now let's use this definition, and try to establish for
example, some some upper bound on some functions.
And I will take as an example, f of n equal
1,000 plus 17n plus 2n squared.
And I want to show, as I said before when we were talking about the growth of
functions, that we can ignore the lower terms, and
we can ignore multiplicative constant factors.
I can ignore the 2, 17 and thousand, and so on.
And just say that this is going to grow in the order of n squared.
Where does, where does that come from?
Or am I just doing some sort of tricks?
If we come actually now and understand the definition of big O, we will see that in
fact, F of N, is O of G of N, where OG of N is n squared.
How do I get that f of n is, is O of G of N in this case?
We have to go back to the definition.
The definition says that if to claim that f of n is O of g of n,
I have to find two constants, C greater than 0, n 0 greater than or equal to 0.
Such that this is going to be bounded from above by
C times n squared for every n greater than or equal to n 0.
Now, we can guess some of these numbers, but
also we can go about it a bit more systematically.
Remember that
I'm interested in finding f of n is smaller than or equal to something.
All right?
So, the way I would do it, is that f of n is 1000 plus 17n plus 2n squared.
Now I want it to be smaller than something, right?
What is it smaller than?
Of course it is smaller than 2 to the n for many cases and
so on, but since I am going after n squared, I have to keep n squared in mind.
Right?
If I want to establish something that is O of n squared, I have to keep n
squared in mind and I have to get to something that looks like n squared.
I can say that 1,000 is smaller than or equal to 1,000 n squared.
If you take a number, and you'll now multiply it by a positive number,
of course this is smaller than or equal to this, right?
1000 is smaller than, or equal to 1000n squared, except if n equals 0.
Right?
If n equals 0, 1000 is not smaller than or equal to 1000 time n squared.
If n is 1, 1000 is smaller than or equal to 1000 n squared.
If n is 2, this is where this starts growing faster now.
1000 is much smaller than 1000 times 4, and so on.
So for this to hold, I have to start now putting conditions that n has to
be greater than or equal to 1.
Okay?
For me to make this jump that 1000 is smaller than or equal to 1000 n squared,
I have to have this assumption that n is greater than or equal to 1.
So this is smaller than this, plus 17 n,
17 n also is smaller than 17 n squared.
Okay?
Again, 17 n is smaller than 17 n squared because we have n is, n squared is,
is larger than n, and we can make this assumption, yeah, okay?
In this case, even if n is 0, 17 n is smaller than or
equal to 17 times n squared.
Okay?
So we don't have to modify this condition.
This is still going to hold.
I cannot lower it to 0 because it's going to violate this one.
Plus, 2n squared is smaller than or equal to 2n squared, okay?
Now, you might be asking now, but also 1000 is smaller than or equal to 1000n.
Or you might be asking, but 1000 is also smaller than or
equal to 1000 n the fifth, which is true for this condition.
Why did I, you choose N squared?
Again, keep in mind that I'm trying to establish that this is O of n squared.
So I would like to see N squared here, because once I see N squared here,
now this is a term that's easy for
me to manipulate, because now I can add these numbers here and
say 1,000 plus 17 plus 2 is 1,019 n squared.
Right?
So now I have this function, that is guaranteed to be smaller than or
equal to 1,019 n squared, for every n greater than or equal to 1.
Again, I can basically say that 1,000, plus 17 n plus 2n squared,
is smaller than or equal to 1,019 n squared for
every N greater than or equal to 1, okay?
Now, when I look at something like this, and look at this kind of definition,
now you can start seeing some sort of correspondence.
This, is F of N.
This is G of N.
So what do I have here?
F of N, now I have F of N is smaller
than 1019 times G of N for every N greater than 1.
The last thing I need to do, is try to find C and
n 0, but they actually jump at you.
Once you look at it like this, because, if you look at this, just call this C,
call this n 0.
We have to make sure that this if I call them C and
n 0, that they satisfy the, the properties of the finishin.
The finishin says that C is greater, has to be greater than 0.
It is greater than zero.
For example, if doing this exercise I got to minus three here,
I couldn't call minus three C, because C has to be greater than 0.
I call this n 0.
N 0 has to be greater than or equal to zero.
1 is greater than or equal to 0.
So I am done.
Okay? I am done.
I have showed that I can find C,
I can find an n 0, such that this function is smaller than or
equal to C times n squared, for every n greater than or equal to n0.
The C is 1,019, then n0 equal 1.
So, this is what I meant in the beginning when I said that we
will be doing rigorous mathematical reasoning rather than proving.
I'm not saying that this is a full proof that this is a,
that this is true, but we can argue, or this is, this is not the technique for
proving formally that f of n is g of n, and it might not work for many cases.
But it's easier for me to reason about how why this is smaller than C times this,
by using this kind of technique, okay?
So here, I just showed a specific value of C, a specific value of C of, of n0,
such that I can bound f of n from above by C times n squared, for
every value of n greater than or equal to 1.
And this is how we would basically do this kind of, of of exercise.
Now, if this was omega, for example if I said,
f of n is omega of, of n squared, we can do a similar exercise.
But then, we shouldn't be focusing on showing it is smaller than or
equal to something else.
We should be focusing on greater than or equal to something else, okay?
And the last thing I want to say is that, if you want to show a function f is theta
of g of n, and you have showed that the function f f is O of g of n,
and the function f is omega of g of n, in fact, you are done.
Because it's easy to, to see that if f
is f of n, is big O of g of n,
and f of n is omega of g of n.
You can conclude from these two that f
of n is theta of g of n, okay?
In fact, you can also go this way.
If you prove that f of n is theta of g of n, you have, in fact, proved that f
of n is O of g of n, and at the same time, f of n is omega of g of n.
Explore our Catalog
Join for free and get personalized recommendations, updates and offers.
Coursera provides universal access to the world’s best education,
partnering with top universities and organizations to offer courses online.
© 2019 Coursera Inc. All rights reserved.
