Distributions

Hello and welcome to my ongoing video course about Distributions, already consisting of 20 videos. Alongside the videos, I provide helpful text explanations. To test your knowledge, take the quizzes and refer to the PDF versions of the lessons if needed. If you have any questions, feel free to ask in the community forum. Now, without further ado, let’s get started!

Part 1 - Motivation and Delta Function

Distributions is a video series I started for everyone who is interested in calculating with generalised functions in the realm of partial differential equations or in physics. We will start with an overview and quickly define the new concepts in a rigorous way. It will be helpful if you have some knowledge in the basics of Linear Algebra and Real Analysis. My video courses can help you there if you want to refresh your memory. Use them while watching the videos about distributions whenever you don’t understand a term we use. On the other hand, if you are interested in the topics about distributions, you might also be interested in my Functional Analysis and Measure Theory series. However, they are not necessarily needed to understand this course but still some videos could help you for getting more deep insights into the field of distributions.


YouTube YouTube Vimeo (ad-free)

PDF PDF (print) PDF Quiz Ask Help


Content of the video:

00:00 Intro
00:16 Analysis
01:30 Historical Example
03:55 Delta is not an ordinary function
06:10 Visualisation
8:00 Delta Function
9:20 Outro


Part 2 - Test Functions

After this short historic overview, we can go into the mathematical definitions. We start by defining the so-called test functions:


YouTube YouTube Vimeo (ad-free)

PDF PDF (print) PDF Quiz Ask Help


 

Part 3 - Convergence of Test Functions

Now you know the space of test functions as a set. However, they have also some important properties, we should discuss. One that is immediately used is a notion of convergence.


YouTube YouTube Vimeo (ad-free)

PDF PDF (print) PDF Quiz Ask Help


 

Part 4 - Space of Distributions

At this point, we are ready to define the important space of distributions. It turns out that this is not so complicated when you are already familiar with linear maps.


YouTube YouTube Vimeo (ad-free)

PDF PDF (print) PDF Quiz Ask Help


 

Part 5 - Regular Distributions

Next, we look at more examples for distributions. Most importantly, we will define so-called regular distributions. This will be the starting point for a lot of definitions that will follow.


YouTube YouTube Vimeo (ad-free)

PDF PDF (print) PDF Quiz Ask Help


 

Part 6 - The Delta Distribution Is Not Regular

We have already introduced the famous delta distribution. Now, we can show that this is not a regular distribution.


YouTube YouTube Vimeo (ad-free)

PDF PDF (print) PDF Quiz Ask Help


 

Part 7 - Distributions Form a Vector Space

Let’s go one step back and look at the space we have defined. Distributions are linear functionals with an additional property: the particular continuity defined above. However, the space still, very naturally, forms a vector space. This means that we can add and scale distributions.


YouTube YouTube Vimeo (ad-free)

PDF PDF (print) PDF Quiz Ask Help


 

Part 8 - Multiplication with Smooth Functions

In the same way, we could define the multiplication of distributions. However, it turns out that, in general, the product of two distributions cannot be defined in the useful manner. Therefore, we restrict ourselves to the multiplication of a distribution with a smooth function.


YouTube YouTube Vimeo (ad-free)

PDF PDF (print) PDF Quiz Ask Help


 

Part 9 - Coordinate Transformation

Since we are working in the n-dimensional space $ \mathbb{R}^n $, coordinate transformations are very natural. The simplest ones, can be described by invertible linear maps. Let’s see how distributions should change under such transformations.


YouTube YouTube Vimeo (ad-free)

PDF PDF (print) PDF Quiz Ask Help


 

Part 10 - Distributional Derivative

This is one of the most important definitions for distributions: we will extend the classical definition of a derivative, which only holds for smooth functions, to the whole realm of distributions, which don’t have to be smooth at all. It turns out that this extension works for every distributions such that a lot of technical problems for differential equations are already solved with this.


YouTube YouTube Vimeo (ad-free)

PDF PDF (print) PDF Quiz Ask Help


 

Part 11 - Fundamental Solution

The next video is interesting for everyone who already learnt a little bit about partial differential equations. There one usually starts with an important example: the Laplace equation. For this is turns out that one has a so-called fundamental solution that helps constructing solutions for the inhomogeneous equation. By using distributions, in particular by using the delta distribution, one can put a rigourous meaning to this concept.


YouTube YouTube Vimeo (ad-free)

PDF PDF (print) PDF Quiz Ask Help


 

Part 12 - Finite-Order Distributions

The following video just wants to show a nice concept of some special distributions is not really needed for the rest of the series. However, it’s nice to see that some distributions coincide with ordinary measures as one learn about them in the Measure Theory series.


YouTube YouTube Vimeo (ad-free)

PDF PDF (print) PDF Quiz Ask Help


 

Part 13 - Convolution

In this video, we will first talk about the convolution as an operation for functions. Then we will also see how we can generalize this to get a convolution between test functions and distributions.


YouTube YouTube Vimeo (ad-free)

PDF PDF (print) PDF Quiz Ask Help



 

Part 14 - Convolution to Solve Partial Differential Equations

Now we go deeper in the study of the convolution. First let’s show that the delta distribution acts like a neutral element there.


YouTube YouTube Vimeo (ad-free)

PDF PDF (print) PDF Ask Help


 

Part 15 - Support for Distributions

In order to generalize the definition of the convolutions even more, we first have to talk about the support for a distribution $T \in \mathcal{D}^\prime(\mathbb{R}^n)$. In short, it should describe at which points the distribution does not vanish. However, this requires some explanations about what this actually means because, in general, one cannot how a distribution acts on single points.


YouTube YouTube Vimeo (ad-free)

PDF PDF (print) PDF Ask Help



 

Part 16 - Distributions with Compact Support

Let’s define a subset of special distributions, denoted by $\mathcal{E}^\prime(\mathbb{R}^n)$. These are the distributions of compact support. It turns out that one can extend them to a bigger domain of definition, which we will also discuss here.


YouTube YouTube Vimeo (ad-free)

PDF PDF (print) PDF Ask Help


 

Part 17 - Convolution with Distributions of Compact Support

By using the special distributions of compact support, one can extend the definition of the convolution even more such that the convolution between two distributions makes sense.


YouTube YouTube Vimeo (ad-free)

PDF PDF (print) PDF Ask Help


 

Part 18 - Fundamental Lemma of the Calculus of Variations

Before we continue, you should notice that all the notions in this series can also be defined for an open set $\Omega \subseteq \mathbb{R}^n$ instead of the whole space $\mathbb{R}^n$. For example, a test function $\mathcal{D}(\Omega)$ is a $C^\infty$-function that has a compact support and this one has lie completely in $\Omega$. In the same sense, we can also define distributions and locally integrable functions on $\Omega$. After this, we can also formulate the general version of the so-called fundamental lemma of the calculus of variations. This one states that testing a locally integrable function with test functions is enough to conclude that it has to be the zero function.


Early Access (for Steady members) YouTube Early YouTube Dark Early Vimeo (ad-free)

PDF PDF (print) PDF Ask Help


 

Part 19 - Proof of Fundamental Lemma of the Calculus of Variations

In order to prove the statement from the last video, we have to know some measure and integration theory. In particular, we can use the fact that the test functions lie dense in space of integrable functions, which is shown in my Multidimensional integration course. Combining the knowledge from there gives us the full proof:


Early Access (for Steady members) YouTube Early YouTube Dark Early Vimeo (ad-free)

PDF PDF (print) PDF Ask Help



 

Part 20 - Sobolev Spaces

We already know that the distributional partial derivative of a distribution always exist. Now, it can definitely happen that the result is a regular distribution, which means it can be identified with a locally integrable function in $L^1_{\mathrm{loc}}(\Omega)$. If this is the case, we actually call it a weak derivative. This idea gives rise to the so-called Sobolev spaces, which combine this weak differentiability with the $p$-integrability of the $L^p$-spaces.


Early Access (for Steady members) YouTube Early YouTube Dark Early Vimeo (ad-free)

PDF PDF (print) PDF Ask Help



 


Connections to other courses


Summary of the course Distributions


Ad-free version available:

Click to watch the series on Vimeo.


Do you search for another mathematical topic?