You are on page 1of 9

Vector basics

vector: indexed list of numbers

3 6

elements

indices

x1

x2

x3

x4

if three elements or fewer, can draw as arrow


x3

1 2 1

2
1
x1

Monday, January 13, 14

x2

Vector basics
if more than three elements, hard to draw, but can still
think of as point in high-dimensional space

x = x1

Monday, January 13, 14

x2

x3

x4 : arbitrary point in 4-D space

Vector basics
simple operations
dot product: x y = x1 y1 + x2 y2 + ... + xn yn
takes two n-D vectors and outputs scalar
norm: kxk =

2
x1

2
x2

+ ... + x2n

takes one n-D vector and outputs scalar

Monday, January 13, 14

Vector basics
simple operations
x
=
finding a unit vector: x
kxk

takes in n-D vector, outputs n-D vector with unit


norm

Monday, January 13, 14

Vector basics
tells you how
the dot product of y with unit vector x

much of y lies in the direction of x


the projection of y onto the direction
we often call y x

defined by x
= 0 , we say that y is orthogonal to x

if y x

in two or three dimensions, orthogonal is synonymous


with perpendicular

Monday, January 13, 14

Functions are vectors


almost everything you can do with vectors you can also
do with functions
to see the equivalence, remember that a function is
essentially an indexed list of numbers also
when talking about functions we use the word argument
instead of index and function value instead of element

Monday, January 13, 14

Functions are vectors


example: writing x(t) = 2t + 3 like a vector
function value (element)
x(t) :
t:

...

7.2

7.0

6.8

...

2.8 3.0 3.2

5.1

5.0

4.9

0.1 0.0

0.1

argument (index)
in the real world (pun intended) the values that t can
take on are infinitely close together, but intuitively its
easier to think of them as having some small spacing
Monday, January 13, 14

...

Functions are vectors


since functions are basically really high dimensional
vectors, we can define vector-like operations on them
the dot product becomes the inner product:
hx(t)|y(t)i =

x(t)y(t)dt
a

takes in two functions and outputs a scalar


is generalization of multiplying elements pairwise and
adding them all up
two functions are orthogonal if their inner product is
zero
Monday, January 13, 14

Functions are vectors


function norm: kx(t)k =

x(t)2 dt
a

takes in function and outputs scalar


x(t)
finding a unit function: x(t) =
kx(t)k

takes in function and outputs function with unit norm

projections: hy(t)|x(t)i =

y(t)
x(t)dt
a

we say this is the projection of y(t) onto the direction


defined by x(t)
Monday, January 13, 14