Learning MATLAB by Doing MATLAB

Learning MATLAB by doing MATLAB
December 10, 2005

Just type in the following commands and watch the output.
1. Variables, Vectors, Matrices

>a=7
>b=[1,2,3]
>c=[1+2,3,3]
>d=[7 7 2]
>e=[7 a 2]
>f=[1;2;3;4]
>g=f(2)
>E=[1 2 3;2 1 3]
>h=E(1,2)
>E
>F(3,4)=7
>F(4,3)=2
>F(1,2)=3
>F(1:2,3:4)=[1 3;2 7]
>whos
>clear a b
>whose
>clear
>help clear
>A=[1 2];
>A
>pi
>A=eye(3)
>b=[1 2 3]
>B=diag(b)
>C=diag([1 7 8])
>D=diag([1;7;8])
>E=ones(4)
>F=ones(2,3)
>G=zeros(4)
>H=zeros(2,3)
>I=[A B;zeros(3) A]
>C
>C
>w(3)=5
>x=0:1/3:2
>for i=1:10
y(i)=2*i
end
>for i=1:10
z(i)=2*i;
end
>z
This
a is interpreted as a scalar (or 1 1 matrix)

after comma: new element in the same row, therefore we have b R1,3
spaces have the same meaning as commas
semicolon: starts a new row, therefore we have f R4,1
accesses the second element of the column vector f
MATLAB distinguishes between lower and upper cases
accesses the (1, 2) element of E
MATLAB currently considers F to be a 3 4-Matrix.
Void elements are set to zero.
Now we need a fourth row!
Sets the (1, 2) element of F to 3.
1 : 2 means row 1 to row 2; 3 : 4 means column 3 to column 4
statistics of used variables
deletes variables a and b
deletes all variables
help for the command clear
concluding the line with a semicolon suppresses the output
3 3 identity matrix
diagonal matrix
geht auch mit Spaltenvektoren
4 4 matrix of all ones
2 3 matrix of all ones
block matrix
transpose of C
MATLAB considers w to be a row vector. The third element of w is set to 5.
row vector with entries from 0 to 2 in steps of 1/3
i runs from 1 to 10. (MATLAB waits for the final end kommt!)
the index i = 0 is not possible here, since vectors have no 0th element!
is partly based on lecture notes by Christian Mehl and Andreas Steinbrecher, both TU Berlin.
2. Simple Operations
>clear
>A=[1 2 3;2 1 0]
>B=[2 2;1 0;0 1]
>C=[0 1 0;5 1 3]
>size(A)
>[m,n]=size(A)
>m=size(A,1)
>b=[2 1 3]
>x=[2;1;3]
>A*B
>A*C
>A
>C
>A*C
>diag(A*C)
>whos
>D=A+C
>E=A+B
>E=A-B
>g=A*x
>g=A*b
>A
>b
>B
>f=b*B
>C=[1 2 3]
Returns number of rows and columns of A in a row vector.

One way to access these numbers individually.
Another way.
matrix multiplications
error, dimensions do not fit!
A CT
returns the diagonal of a matrix
ans (for answer) is the is the unnamed output
matrix addition
error, dimensions do not fit!
E = A BT
matrix times vector
error!
row vector times matrix
3. Matrix Manipulations
>clear
>A=[1 2;3 4]
>A(3,2)=7
>A(1:2,2)
>A(3,1:2)
>B(3:4,3)=[5;6]
Adds a third row!

(1:2,2): 1st to 2nd element of column 2
(3,1:2): 1st to 2nd element of row 3
MATLAB creates a matrix B with the 3rd and 4th elements
of column 3 being 5 and 6, respectively. All other entries are zero.
>C(4:5,4)=A(1:2,2)
>B(:,3)
3rd column of B
>d=C(1,:)
1st row of C
>E=[1 2 3;4 5 6;7 8 9;10 11 12]
>E(1:2:4,3)
(1:2:4,3): Picks every second element in column 3 from 1 to 4.
>F=[1 2 3 4 5;6 7 8 9 10;11 12 13 14 15]
>G=F(1:2:3,1:2:5)
Picks every second element in columns 1,3,5 from 1 to 3.
>G=F(1:2:end,1:2:end)
Alternative notation if you forgot the dimensions.
>H=[1 3;9 11]
>H^(-1)
The inverse.
>inv(H)
Also the inverse.
>H\A(1:2,1:2)
Always use the backslash for computing H 1 A. Dont use inv(H)*A!
>det(H)
The determinant.
>b=[99 100 101]
>F(1,1:3)=b
>A
>A(1,1:3)=b
Elements (2, 3) and (3, 3) are added.
4. Subprograms, m Files
For the following subprograms you have to create files with the name of the function appended by .m, for example
test1.m. (You may use any editor to this, MATLAB comes with a built-in editor which also offers debugging.)
The m files have to be in the current working directory. Try help pwd and help cd for more infos about working
directories.
% This is my first m file.
% You can add comments behind %.
% It is called test1.m and computes the sum of all elements of a matrix A.
function y=test1(A)
[m,n]=size(A);
y=0;
% Init
for i=1:m
% i runs from 1 to m
for j=1:n
% j runs from 1 to n
y=y+A(i,j);
end
Save under test1.m.
end
>clear
>help test1
>A=[1 2;3 4]
>s=test1(A)
% The program test2.m computes the sum and the product of all elements of a matrix A,
% as well as the trace if the matrix is square.
function [y,p,t]=test2(A)
[m,n]=size(A);
summe=0;
p=1;
t=0;
for i=1:m
if (m==n)
% if m = n, then ...
t=t+A(i,i);
end
for j=1:n
y=y+A(i,j);
p=p*A(i,j);
end
end
Save under test2.m.
>clear, help test2

>A=[1 2;3 4]
>test2(A)
>[su,pr,sp]=test2(A);
>su,pr,sp
>B=[1 2 3;4 5 6]
>[su,pr,sp]=test2(B)
Only the first output argument is displayed.
MATLAB functions can have several input variables (e.g., function ausgabe=test3(A,B,C)) or require no input/output at all (e.g., function []=test4() or you just omit the whole function declaration, call with test4).
Besides the for loop, there is also a while loop. See help while.
5. Graphics
>for i=1:10, x(i)=i/10;
>plot(x,y)
>plot(x,z)
>clf
>plot(x,y)
>hold on
>plot(x,z)
>plot(x,2*z,r)
>plot(x,y+z,g*)
>hold off
>plot(x,y-z,k+)
>help plot
>title(My plot)
>xlabel(x axis)
>ylabel(y axis)
>axis([0,20,-5,50])
>box
>grid
>clf
>subplot(3,2,1)
>plot(x,y)
>subplot(3,2,2)
>plot(x,z,k)
>subplot(3,2,5)
>plot(x,z+y,mo)
>hold on
>plot(x,z,k)
>subplot(3,2,1)
>plot(x,z,k)
>subplot(3,2,4)
>title(leer)
>subplot(3,1,2)
>plot(y)
>orient tall
>help orient
>print -dps test1.ps
>help print
y(i)=x(i)^2; z(i)=sqrt(x(i)); end
The plot has 3 2 = 6 subplots. The 1st subplot is active.
We have only 3 subplots, one in each row. The 2nd is now active.
Creates a pos file test1.ps of this plot (more options in the menu File/Save As..)
Turn on Tools/Edit Plots and right click to change properties of the plots. For inclusion of plots in presentations and
papers it is often a good idea to increase the font sizes and thicken the lines.
6. Other useful commands
load
save
blkdiag
kron
A(:)
.* ./
chol
cond
eig
hess
norm
qr
schur
svd
load variables from files

save variables in files
create block diagonal matrices
Kronecker product
vec(A)
elementwise multiplication and division of two matrices
Cholesky decomposition of matrices
matrix condition number
eigenvalues of matrices and matrix pencils
Hessenberg form of matrices
matrix norm
QR decomposition of matrices
Schur form of matrices
singular values of matrices
7. Using the Control Toolbox

First, let us consider a simple example for demonstrating the basic functionality.
>A = [0 1; -1/4 -1], B = [ 0; 1/2 ]
>C = [0 1], D = 0
Create the system matrices of a simple system.
>sys1 = ss(A,B,C,D)
The object sys1 contains this system in state space form.
>sys1.a
>eig(sys1)
poles of the system
>tf1 = tf(sys1)
form transfer function
>eig(tf1)
surprise, surprise
>zero(tf1)
>ss(tf1)
Can you tell the difference?
>bode(sys1)
Try to guess the H norm!
>norm(sys1,inf)
>20*log10(ans)
Converted to dB.
>norm(sys1,2)
>svd(ctr(sys))
Kalman test for controllability
>svd(obsv(sys))
Kalman test for observability
>A = [4 3; -9/2 -7/2], B = [ 1; -1 ],
>C = [3 2], D = 0,
Create the system matrices of another system.
>sys2 = ss(A,B,C,D)
>eig(sys2)
>big = sys1 + sys2
Convenient way to form interconnected systems.
>zero(big), eig(big)
>norm(big,inf)
>norm(big,2)
>svd(obsv(big))
>[Ao,Bo,Co,Qo,ko] = obsvf(big.a,big.b,big.c)
>svd(ctrb(big))
>[Ac,Bc,Cc,Qc,kc] = ctrbf(big.a,big.b,big.c)
>bigmin = ss(big,min)
>zero(bigmin), eig(bigmin)
>sys1.d = 1
Make first system invertible.
>tf(sys1)
>sysi = inv(sys1)
Form inverse of system.
>tf(sysi)
>s = tf(s)
Convenient for creating transfer functions directly.
>H = (s-1)*(s+4)/(s^2+2*s+10)
Type help control for a list of available functions in the control toolbox.
Now, let us consider a bigger example.

1. Create a working directory if you have not done so far. Download the benchmark examples available from
http://web.math.hr/kressner/nlacontrol/ into this directory.
2. Store the system matrices of the benchmark example 1.9, which represents an aeroelastic model of parts of a
Boeing B-767 airplane, in a system object in state space form (you can ignore the identity matrix E).
3. Investigate the stability of this system by computing its poles.
4. Have a look at the coefficients of the transfer function.
5. Form the controllability matrix and plot its singular values with semilogy.
6. Test controllability/obserability using the Hautus test.
7. Determine the number of controllable poles using ctrbf. Investigate the structure of the state matrix in controllability form using spy(abs(Ac)>0.00000001).
8. Determine the number of observable poles using obsvf.

Learning MATLAB by Doing MATLAB

Uploaded by

Copyright:

Available Formats

Learning MATLAB by Doing MATLAB

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Learning MATLAB by Doing MATLAB

Uploaded by

Copyright:

Available Formats

Learning MATLAB by doing MATLAB

December 10, 2005

1. Variables, Vectors, Matrices

a is interpreted as a scalar (or 1 1 matrix)

Returns number of rows and columns of A in a row vector.

row vector times matrix

Adds a third row!

>clear, help test2

Only the first output argument is displayed.

y(i)=x(i)^2; z(i)=sqrt(x(i)); end

The plot has 3 2 = 6 subplots. The 1st subplot is active.

load variables from files

7. Using the Control Toolbox

Now, let us consider a bigger example.

You might also like