How Do You Know if a Vector Is in the Span of S

"Mathematics requires a small dose, not of genius, but of an imaginative freedom which, in a larger dose, would be insanity."

$\qquad$ — Angus G. Rodgers

In the terminal chapter, along with the ideas of vector addition and scalar multiplication, I described vector coordinates, where there'southward this back-and-along between pairs of numbers and two-dimensional vectors.

At present I imagine that vector coordinates were already familiar to many of you lot, just at that place's another interesting way to think most these coordinates, which is primal to linear algebra. When you lot have a pair of numbers meant to draw a vector, like $(3, -2)$ , I desire y'all to think of each coordinate as a scalar, meaning think about how each ane stretches or squishes vectors.

In the $xy$ -coordinate system, there are two special vectors. The ane pointing to the right with length $1$ , commonly called "i hat" $\chapeau i$ or "the unit vector in the x-direction". The other one is pointing straight up with length $ane$ , commonly chosen "j hat" $\chapeau j$ or "the unit of measurement vector in the y-direction". Now, think of the x-coordinate equally a scalar that scales $\hat i$ , stretching it by a cistron of $3$ , and the y-coordinate equally a scalar that scales $\hat j$ , flipping it and stretching it past a factor of $ii$ .

In this sense, the vector that these coordinates depict is the sum of two scaled vectors. This thought of adding together two scaled vectors is a surprisingly of import concept. Those 2 vectors $\chapeau i$ and $\hat j$ have a special name: Together they are chosen the "basis" of the coordinate arrangement. What this means is that when y'all think about coordinates every bit scalars, the footing vectors are what those scalars actually scale.

In that location'south as well a more than technical definition of basis, merely we'll get to that subsequently. Framing our familiar coordinate system in terms of these two special basis vectors raises an interesting and subtle point: We could choose a different pair of basis vectors and become a perfectly reasonable new coordinate system.

Choosing Different Basis Vectors

For example, take some vector pointing up and to the right, along with a vector pointing down and to the correct.

In the above figure, $\overrightarrow{\mathbf{v}}=\brainstorm{bmatrix}1\\ 2\end{bmatrix}$ and $\overrightarrow{\mathbf{westward}}=\begin{bmatrix}iii\\ -1\end{bmatrix}$ .

What values of the scalars $\alpha$ and $\beta$ satisfy the following equation? $\alpha\vec{\mathbf{v}}+\beta\vec{\mathbf{w}}=v\hat i -\frac12\lid j$

$\blastoff=1\ \beta=0.5$

$\alpha=0.5\ \beta=1.five$

$\alpha=-1\ \beta=-1.5$

$\alpha=0.5\ \beta=-1.5$

Nosotros have a new pair of basis vectors $\overrightarrow{\mathbf{5}}$ and $\overrightarrow{\mathbf{due west}}$ . Take a moment to think about all the different vectors you tin can go past choosing two scalars, using each to calibration one of the vectors, and so adding them. Which two-dimensional vectors tin can you reach past altering your pick of scalars?

The answer is that yous can describe every possible two-dimensional vector this way, and I think information technology's a practiced puzzle to contemplate why. A new pair of basis vectors like this still gives you lot a way to get back and forth between pairs of numbers and two-dimensional vectors, only the association is definitely different from the version you get using the standard basis of $\chapeau i$ and $\hat j$ .

I'll go into much more particular on this point in a later chapter, describing the relationship betwixt different coordinate systems, simply for now I but desire you to appreciate that whatever way to depict vectors numerically depends on your choice of footing vectors.

Linear Combinations

Any time you're scaling two vectors and adding them like this, it'southward called a "linear combination" of those two vectors. Where does the word "linear" come up from here? What does this have to do with lines? Well, when you multiply a scalar by a vector, it changes the magnitude of that vector. Multiplying every real number by the vector produces an infinite line that passes through the origin and the point divers by the vector.

So a linear combination of two vectors is a method of combining these ii lines. For most pairs of vectors, if yous allow both scalars range freely and consider every possible vector yous could go, you will be able to accomplish every possible betoken on the plane. Every two-dimensional vector is within your grasp.

Nonetheless, if your two original vectors happen to line up, the lines produced past the scalar multiplication volition be the same line, and then adding them together can't yield a vector outside of that line. In that location'due south a third possibility besides: Both your vectors could be the 0 vector, in which case you lot'll just exist stuck at the origin.

Span

The set of all possible vectors y'all can achieve with linear combinations of a given pair of vectors is called the "bridge" of those two vectors. Restating what we only saw in this lingo, the span of most pairs of 2D vectors is all vectors in 2D space, only when they line up, their bridge is all vectors whose tip sit on a certain line.

Remember how I said linear algebra revolves around vector addition and scalar multiplication? The span of two vectors is basically a way of asking what are all the possible vectors you tin can reach using these two by only using those fundamental operations of vector improver and scalar multiplication.

What is the span of $\begin{bmatrix}3\\ -2\end{bmatrix}$ and $\brainstorm{bmatrix}-6\\ 4\end{bmatrix}$ ?

$\mathbf{0}$

$\mathbb{R}$

$\mathbb{R}^two$

Vectors vs Points

This is a good time to talk virtually how people commonly think most vectors as points. Information technology gets very crowded to recall about a whole drove of vectors sitting on a line, and more crowded nevertheless to think about all two-dimensional vectors all at once, filling up the aeroplane.

Then when dealing with collections of vectors similar this, information technology'due south common to represent each one but with a point in space, the betoken at the tip of this vector. That way, if you want to think about every possible vector whose tip sits on a certain line, just call back about that line itself.

Likewise, to think about all possible 2-dimensional vectors, conceptualize each one every bit the point where its tip sits. Then to think about all of them all at once, you can just recall well-nigh the infinite flat sheet that is ii-dimensional space, leaving the arrows out of information technology.

In general, if yous're thinking of a vector on its own, think of information technology as an arrow, and if y'all're thinking of a drove, it's convenient to remember of them as points.

Span in 3D

The thought of span gets more than interesting if nosotros start thinking about vectors in three-dimensional space. For example, if yous take 2 vectors in iii-dimensional space that are non pointing in the same management, what does it hateful to have their span?

Well, their span is the collection of all possible linear combinations of those two vectors, meaning all possible vectors you become past scaling each of the two y'all get-go with in some way, then adding them together. You can imagine turning two knobs to modify the two scalars defining the linear combination, calculation the scaled vectors and following the tip of the resulting vector. That tip traces out some kind of flat canvass cutting through the origin of iii-dimensional space.

$$\color{green}\overrightarrow{\mathbf{ten}} \color{black}= \color{blood-red}a\overrightarrow{\mathbf{v}} \color{blackness}+ \color{purple}b\overrightarrow{\mathbf{w}}$$

Which vectors in 3D space are not in this span?

The ready of all possible vectors whose tips sit on this flat sheet is the span of your 2 vectors. Any vector which does not prevarication on the plane is not in the span.

So, what happens if you lot add on a third vector, and consider the span of all three of those guys? A linear combination of three vectors is defined pretty much the same way equally for ii: Choose three scalars, apply them to calibration each of your vectors, and then add them all together. And again, the span of these vectors is the set of all possible linear combinations.

$$\color{greenish}\overrightarrow{\mathbf{x}} \color{black}= \color{red}a\overrightarrow{\mathbf{v}} \colour{black}+ \color{regal}b\overrightarrow{\mathbf{west}} \color{black}+ \color{blue}c\overrightarrow{\mathbf{u}}$$

Two things could happen. The get-go possibility is if your third vector happens to be sitting on the span of the starting time 2. Then the span doesn't change, you're sort of trapped on that same flat sheet. In other words, calculation a scaled version of the third vector to linear combinations of the first ii doesn't requite yous admission to any new vectors. This means the third vector can also be expressed equally a linear combination of the other two:

$$\colour{blue}\overrightarrow{\mathbf{u}} \color{black}= \color{red}a\overrightarrow{\mathbf{v}} \color{black}+ \color{royal}b\overrightarrow{\mathbf{westward}}$$

There is some other possibility though, if y'all merely randomly choose a third vector, it'due south almost certainly non sitting on the span of the first. Since it'due south then pointing in a split up direction, it unlocks access to every possible three-dimensional vector! The mode I like to retrieve most this is that as yous scale the new third vector, it moves around the bridge of the first two to sweep it through all of infinite.

It's kind of like you're making full use of the three freely-irresolute scalars that you have at your disposal to access the full three dimensions of infinite.

In the instance where the third vector was sitting on the span of the offset two, or the example where 2 vectors happen to line up, we desire some terminology to describe the fact that at to the lowest degree 1 of these vectors is redundant, not adding annihilation to our span. Whenever this happens, where you accept multiple vectors, and y'all could remove one without reducing their span, the relevant terminology is to say they are "linearly dependent".

Another style of phrasing this would exist to say that one of the vectors can be expressed as a linear combination of the others. That is, it'due south already in the span of the other two. On the other hand, if each vector really does add another dimension to the span, they are said to be "linearly independent".

$$\brainstorm{align*} \text{Linearly Dependent: }\quad \color{bluish}\overrightarrow{\mathbf{u}} \color{blackness}&= \color{ruby}a\overrightarrow{\mathbf{5}} \color{black}+ \colour{regal}b\overrightarrow{\mathbf{w}} \quad\colour{blackness}\text{ for some }\color{red}a\color{black}\text{ and }\color{royal}b \\ \text{Linearly Independent: }\quad \color{blue}\overrightarrow{\mathbf{u}} \color{blackness}&\neq \color{red}a\overrightarrow{\mathbf{v}} \color{black}+ \color{purple}b\overrightarrow{\mathbf{w}} \quad\color{black}\text{ for all }\color{red}a\color{black}\text{ and }\colour{purple}b \end{marshal*}$$

The technical definition for the "ground" of a space is a set of linearly independent vectors that bridge that space, given how I described a basis earlier, and given your understanding of the words "bridge" and "linearly independent", why does this definition make sense?

Earlier we learned that any pair of vectors could form a new basis every bit long as they didn't line upward. If pair of vectors are linearly independent, their linear combination can span the entire 2d aeroplane, pregnant they can grade the basis for the plane.

flemingmanotty1982.blogspot.com

Source: https://www.3blue1brown.com/lessons/span

Share this post