Scribd
Upload
95 views7 pages

Linear Functions: EE263 Autumn 2015 S. Boyd and S. Lall

The document discusses linear functions and systems of linear equations. It defines a linear function as one where superposition holds, and explains that any linear function can be represented by a matrix multiplication operation f(x) = Ax. It interprets the matrix A as containing gain factors that describe how each input affects each output. The document also discusses how nonlinear functions can be linearized around a point to relate input and output deviations with a linear function defined by the Jacobian matrix.

Uploaded by

Mouli
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
95 views7 pages

Linear Functions: EE263 Autumn 2015 S. Boyd and S. Lall

The document discusses linear functions and systems of linear equations. It defines a linear function as one where superposition holds, and explains that any linear function can be represented by a matrix multiplication operation f(x) = Ax. It interprets the matrix A as containing gain factors that describe how each input affects each output. The document also discusses how nonlinear functions can be linearized around a point to relate input and output deviations with a linear function defined by the Jacobian matrix.

Uploaded by

Mouli
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 7

EE263 Autumn 2015 S. Boyd and S.

Lall

Linear functions

1
Linear equations

consider system of linear equations

y1 = a11 x1 + a12 x2 + · · · + a1n xn


y2 = a21 x1 + a22 x2 + · · · + a2n xn
..
.
ym = am1 x1 + am2 x2 + · · · + amn xn

can be written in matrix form as y = Ax, where

a11 a12 · · ·
     
y1 a1n x1
 y2   a21 a22 · · · a2n   x2 
y= .  A= . x= . 
     
.. .. 
 ..   .. . .   .. 
ym am1 am2 · · · amn xn

2
Linear functions

a function f : Rn −→ Rm is linear if

I f (x + y) = f (x) + f (y), ∀x, y ∈ Rn

I f (αx) = αf (x), ∀x ∈ Rn ∀α ∈ R

i.e., superposition holds

x+y
f (y)
y
x
f (x + y)

f (x)

3
Matrix multiplication function

I consider function f : Rn → Rm given by f (x) = Ax, where A ∈ Rm×n

I matrix multiplication function f is linear

I converse is true: any linear function f : Rn → Rm can be written as f (x) =


Ax for some A ∈ Rm×n

I representation via matrix multiplication is unique: for any linear function f


there is only one matrix A for which f (x) = Ax for all x

I y = Ax is a concrete representation of a generic linear function

4
Interpretations of y = Ax

I y is measurement or observation; x is unknown to be determined

I x is ‘input’ or ‘action’; y is ‘output’ or ‘result’

I y = Ax defines a function or transformation that maps x ∈ Rn into y ∈ Rm

5
Interpretation of aij
n
X
yi = aij xj
j=1

aij is gain factor from jth input (xj ) to ith output (yi )

I ith row of A concerns ith output

I jth column of A concerns jth input

I a27 = 0 means 2nd output (y2 ) doesn’t depend on 7th input (x7 )

I |a31 |  |a3j | for j 6= 1 means y3 depends mainly on x1

I |a52 |  |ai2 | for i 6= 5 means x2 affects mainly y5

I A is lower triangular, i.e., aij = 0 for i < j, means yi only depends on


x1 , . . . , xi

I A is diagonal, i.e., aij = 0 for i 6= j, means ith output depends only on ith
input

more generally, sparsity pattern of A, i.e., list of zero/nonzero entries of A, shows


which xj affect which yi
6
Linearization

I if f : Rn → Rm is differentiable at x0 ∈ Rn , then

x near x0 =⇒ f (x) very near f (x0 ) + Df (x0 )(x − x0 )

where
∂fi
Df (x0 )ij =
∂xj x0
is derivative (Jacobian) matrix

I with y = f (x), y0 = f (x0 ), define input deviation δx := x − x0 , output


deviation δy := y − y0

I then we have δy ≈ Df (x0 )δx

I when deviations are small, they are (approximately) related by a linear function

You might also like