ChE 101: Introduction
to MATLAB
Amy Proctor
January 7, 2012
Downloading MATLAB
MATLAB is available for free for Caltech students and
staff! (However, you need to be on the Caltech network to
use the program.)
To download MATLAB, log into
http://software.caltech.edu and enter your IMSS user
name and password. Click on Start Shopping.''
MATLAB can be found under MathWorks, Inc heading.
Download the latest version of MATLAB.
Using MATLAB off campus
Sign up for VPN at
http://www.imss.caltech.edu/help/vpn-signup-form
Download the Caltech VPN (AnyConnect) client at
https://vpn.caltech.edu/+CSCOE+/logon.html
When you are off the Caltech network, log in before
trying to open MATLAB.
Note: This is not necessary if you download the license file
properly on your computer, but you should sign up for VPN
access on the off chance you need to read a paper or get on
the course website off campus
Main MATLAB windows
Command window: The main window where
MATLAB puts you when you launch the application
Characterized by the >> command prompt)
All commands, including those for running programs
you write, are typed in this window
Workspace pane: Lists all of the variables you have
generated so far, along with their size and type.
Main MATLAB windows
Command history window: All commands typed
on the MATLAB prompt get recorded here (including
those from previous sessions)
Figure window: The output of all graphics commands
typed in the command window goes here
Editor window: Window where you write, edit,
create, and save your own programs in files called m-files
(standard ASCII text files).
Main MATLAB windows
I made some videos to illustrate a few points in
the tutorials. This is the first of them.
Video: http://youtu.be/X40TSdR-sFg
Note that the apostrophe characters
surrounding strings in the font I used for
MATLAB commands/inputs will not copy-paste
properly into MATLAB!
Correct them before executing commands or
scripts!
Useful commands to know
who lists variables currently in the workspace
clear clears the workspace and removes all variables
clear x y z clears only the variables x, y, and z
clear all clears all variables and functions from the
workspace
mlock fun locks the function fun so that clear cannot
remove it
Useful commands to know
munlock fun unlocks the function fun so that clear can
remove it
clc clears the command window and moves the cursor to
the top
cfc clears the figure window
quit and exit both quit MATLAB
Control-C kills the current command execution (good for
aborting infinite loops)
Online help
MATLAB provides extensive online documentation:
http://www.mathworks.com/help/techdoc/
Typing lookfor, help, helpwin, and helpdesk in the
command window all provide online help
Online help
help lists topics on which help is available
helpwin opens the interactive help window
helpdesk opens the web browser-based help facility
help topic provides help on topic
look for string lists help topics containing string
demo runs the demo program
Operations
+ (addition), - (subtraction), * (multiplication), /
(division), and ^ (exponentiation) work as expected for
numbers
exp(x), sin(x), log(x), cos(x), etc. also work as
expected for numbers
Use ; to suppress screen output (not necessary but
excessive printing may slow down program)
Creating and working with arrays
x = [1 2 3] or x = 1:3 creates a row vector with
three elements
y = [4; 5; 6] or y = (4:6) creates a column vector
with three elements
A=[1 2 3; 4 5 6] creates a 2x3 matrix
You can add or subtract two vectors of the same size
using + or -
Creating and working with arrays
To multiply or divide elements of two same-sized
vectors term by term use the array operators .* or ./
For normal matrix multiplication or multiplying a matrix
by a scalar, no dot is needed before *
You can raise each element of a matrix to a power using
the array operator .^
Trigonometric functions and elementary math functions
operate on vectors term-by-term
Creating and working with arrays
linspace(a, b, c) creates a vectors with c linearly
spaced elements between a and c
linspace(a, b) creates a vectors with 100 linearly
spaced elements between a and c
x:y creates a row vector [x x+1...z], where z is the
largest x+c (c is an integer) less than or equal to y
x:y:z creates a row vector [x x+y x+2y ... q], where
q is the largest x+cy (c is an integer) less than or equal
to z
Creating and working with arrays
zeros(n) or ones(n) creates an n-by-n matrix of
zeros or ones
zeros(m, n) or ones(m, n) creates an m-by-n
matrix of zeros or ones
For a vector x, length(x) gives the number of
elements
For a matrix M, length(M) gives the number of
columns and size(M) returns a row vector [(# of
rows) (# of columns)]
Creating and working with arrays
For a vector x, x(i) gives the ith element of the vector
For a matrix M, x(i, j) gives element in row i and
column j
For a vector x, max(x), min(x), mean(x), and
sum(x) all work as expected
For a matrix M, max(M), min(M), mean(M),
and sum(M) all return row vectors with values for
each column of the matrix
Creating and working with arrays
find(X) locates all nonzero elements of an array X and
returns the linear indices of those elements.
Can use a relational expressions like X > n to obtain
all elements greater than n, etc.
MATLAB stores arrays as a column of elements
The index in that column is the linear index j
X(j) gives the element of X with linear index j, and
X(end) gives the last element in the lower right
hand corner
Symbolic computation
To declare x and y symbolic variables, type in syms x y
Now you can write expressions such as f = (x + y)^2
Use expand(f) or simplify(f) to multiply out or
factor algebraic or trigonometric expressions
subs(f, x, 5) will substitute 5 for x and will return
(5 + y)^2
subs(f, [x, y], [1, 2]) will substitute 1 for x and
2 for y and return 9
Symbolic computation
diff(f, x) will return the derivative of f with respect to
x
diff(f, x, n) will return the nth derivative of f with
respect to x
int(f, x) will return the indefinite integral of f with
respect to x
int(f, x, a, b) integrates f with respect to x from a to
b
Symbolic computation
solve(x^4 + 1) and solve( x^4 + 1 ) return the
symbolic array of x values that satisfy x^4 + 1 = 0
Use the double command [ex: double(solve
(x^4+1))] to return an array of numbers
To solve for equality relations, use a string
Example: solve( x^4 + 1 = x^3 )
Specify the variable [ex: solve(a*x^2 + b*x + c, a)]
when you are using multiple symbolic variables
Anonymous functions
Anonymous functions (functions without names),
which are created and referred to by their handles,
can be created by the command f = @(input list)
mathematical expression
Example: g = @(x, y, z) x^2*y/z creates a
function g(x, y, z)
g(1, 2, 3) will evaluate the function
Function files
A function file begins with a function definition line:
function [output variables] = function_name
(input variables)
The function name must be the same as the file name,
and square brackets are not required for a single-output
variable
Comment lines should start with a % sign
All comment lines immediately after a function definition
line be displayed by MATLAB if help is sought on the
function (similar to a docstring in Python)
Function file example
function [sum, average] = myfun(a, b, c)
% Finds the average of three numbers a, b, and % c
sum = a + b + c;
average = sum / 3;
% Note that input variable names given in the %
function definition line are local to the
% function, so other variable names and
% values can be used in the function call
Executing a function file
To execute a function file, we go to the command window (or
use it in a script file) and type something like:
[x y] = myfun(1, 4, 5);
Then the variable x will be defined as the sum of 1, 4, and 5
(10), and the variable y will be the average of 1, 4, and 5
(10/3)
Executing a function file
If you nest functions, you must end each function
function [x, y] = main(a, b, c)
...
function x = nest1(a, b)
...
function y = nest2(c)
...
end
end
end
Executing a function file
Otherwise functions do not require end (although if you use
it for one function in a file, you must use it for all of them)
A function file may contain multiple functions, but all
functions written below the first function will be treated as
subfunctions and will not be available outside of the file
Script files
A script file is an m-file that does not begin with a
function definition line (does not take arguments) and
contains a set of commands
A script file is run by typing the name of the file (without
the .m extension) on the command line
Script files work on global variables (variables currently
present in the workspace) and variables from the script
are left in the workspace
Never name a script file the same as a global variable
since MATLAB first searches the workspace before
looking for functions
For loops
A for loop is used to repeat a statement or group of
statements for a fixed number of times
Unlike Python, MATLAB does not require a : but does require
that you use an end to end each for loop
Example:
% Find the sum of a vector s elements using a loop instead of the
% built-in sum command
x = [1 2 3 7 8 19 35 2];
sum = 0;
for i=1:length(x)
sum = sum + x(i)
end
Relational operators
MATLAB uses the following relational operators, which
compare the elements of arrays A and B elementwise (or
compare scalars A and B):
Less than: A < B or lt(A, B)
Greater than: A > B or gt(A, B)
Less than or equal to: A <= B or le(A, B)
Greater than or equal to: A >= B or ge(A, B)
Equal: A == B or eq(A, B)
Not equal: A ~= B or ne(A, B)
Logical operators
&, |, and ~ are the logical array operators AND, OR,
and NOT in MATLAB, which compare the elements of
arrays A and B elementwise (or compare scalars A and
B)
However, &&, ||, and ~ are used as the logical
operators AND, OR, and NOT in MATLAB loops!
While loops
A while loop is used to execute a statement or group of
statements until the condition specified by while is no longer
satisfied
Does not require a : but does require an end
Example:
% Find all the powers of 2 below 1000
v = 1; num = 1; i =1;
while num < 1000
v = [v; num];
i = i + 1;
num = 2^i
end
v
If loops
if-elseif-else statements can be used in MATLAB.
Requires an end for the if statement only!
Example:
i = [5, 10, 20, 3, 1]; j = [-1, 3, 6, 7, 2];
num = 0;
for k=l:length(i)
if le(i(k),4) && lt(j, 2)
num = num + 1;
elseif ge(j, 6)
num = num + 2;
else
num = num + 3;
end
end
Making 2-D plots
The most basic command for plotting in 2-D is plot(x, y,
style-option )
x and y are vectors containing the x- and y-
coordinates of points for a graph
style-option is an optional argument
Can change color ( y for yellow, m for magenta,
c for cyan, r for red, g for green, b for blue,
w for white, and k for black)
or change line style ( - for solid, -- for dashed, :
for dotted, -. for dash-dot, and none for no line)
Making 2-D plots
or the marker style for individual data points ( +
for plus sign, o for circle, * for asterisk, x for x-
mark, . for point, ^ for triangle, s for square, d
for diamond, etc.)
Style options can be combined!
Example: plot(x, y, b-- ) will give a blue dashed line
(no point markers)
Example: plot(x, y, c*-- ) gives a cyan dashed line
connecting cyan asterisk points
Example: plot(x, y, + ) gives unconnected blue
(default color) points marked by plus signs
Making 2-D plots
The axis command can be used to change the axes
commands after a plot is generated in a script
General form: axis([xmin xmax ymin ymax])
Can control only part of the axes limits and let
MATLAB set the other limits automatically by
specifying the desired limits in the axis command
and using inf (or -inf) as the values of the limits
that you would like to be set automatically
axis( equal ) sets equal scale on both axes
Making 2-D plots
xlabel(Your label here )labels the x-axis
ylabel( Your label here ) labels the y-axis
title( Your title here ) titles the plot
Use legend(string1, string2, ...) to produce a
legend using the text in string1, string2, etc. as labels
See
http://www.mathworks.com/help/techdoc/ref/
legend.html for legend s many optional arguments
Making 2-D plots
Multiple data sets can be plotted in one graph
plot(x1, y1, style-option , x2, y2, style-
option , etc.) will plot y1 vs. x1 with whatever style
option is indicated (can be left out) and plot y2 vs. x2
with whatever style option is indicated (can be left out),
etc.
The hold command can also be used to overlay plots
Useful when you are generating plottable data in a for
or if loop
Making 2-D plots
% For i from 1 to 5, plot the function y = ix if i is even or the function y = x/i if i is
% odd for x from 1 to 10. Plot even i functions in red and odd i functions in green.
% Overlay the plots.
x = 1:0.1:10;
for i=1:5
if mod(i, 2) == 0
y = i * x
plot(x, y, r )
hold on
else
y = 1/i * x
plot(x, y, b )
hold on
end
end
xlabel( x )
ylabel( y )
title( Plot example )
hold off
Ordinary differential equations
MATLAB has several built-in functions to solve ODEs
General form: [t, y] = solver(ode_function, tspan, x0,
options)
solver is the ODE solver of choice
ode_function is the name of a user-defined function
containing ODEs you want to solve
tspan is a vector of the form [t0 tfinal]
x0 is the initial condition
options are created using odeset
Ordinary differential equations
The ODE function must be written as a system/array of
first-order equations, so you may need to introduce new
variables and recast the original equations in terms of first
order ODEs in the new variables
Two of the most common are ode23 and ode45, which
are both nonstiff solvers
A stiff equation is a differential equation for which certain
numerical methods for solving the equation are
numerically unstable
Ordinary differential equations
ode15s is a stiff solver based on a variable order method
you can try if you think your equations are stiff
ode23 and ode45 implement second-/third-order and
fourth-/fifth-order Runge-Kutta methods, respectively
ode23 is generally quicker but less accurate than ode45
As a rule of thumb, use ode45 since it should give you
good results for most problems
odefile is a help file that can help you with the syntax for
your ODE functions
Example 1: x = x + t, x(0) = 0
xdot = @(x, t) x + t;
tspan = [0 10];
x0 = 0;
[t, x] = ode45(xdot, tspan, x0);
plot(t, x);
xlabel( Time )
ylabel( x )
Example 2: Coupled ODEs
options=odeset('RelTol',1e-6);
ode_system = @(t, x)[-x(1)-x(2);x(1)-2*x(2)];
[t,x] = ode45(ode_system, [0 5], [1; 1], options);
plot(t, x(:,1), t, x(:,2))
l = legend('x_1', 'x_2');
set(l, 'location', 'EastOutside')
xlabel('Time')
ylabel('x_1 and x_2')
Solving a system of nonlinear
equations with fsolve
[x, fval] = fsolve(fun, x0, options) (fval and options
are optional arguments) solves the problem fun(x) = 0
Starts at x0, a vector of initial guesses for the solution to
the nonlinear system
Tries to solve the equations described in the user-written
function fun with the optimization options specified in
the structure options (set using optimset)
Returns a vector of improved guesses x and the values of
the objective function fval given the guesses in x
fsolve example
F = @(x)[2*x(1) - x(2) - exp(-x(1));
-x(1) + 2*x(2) - exp(-x(2))];
x0 = [-5; -5];
options=optimset('Display','off');
[x,fval] = fsolve(F,x0,options)
x=
fval =
0.5671
1.0e-06 *
0.5671
-0.4059
-0.4059
Fitting data using the curve fitting
toolbox cftool
Video: http://youtu.be/O2X1lw_bzPY
Curve fitting functions
yy = spline(x, Y, xx) uses a cubic spline interpolation
to find yy, the values of the underlying function Y at the
values of the interpolant xx
Cubic splines fit separate cubic polynomials between
successive data points by matching the slopes and the
curvature of each segment at the data points
beta = nlinfit(X, y, fun, beta0) returns a vector beta
of coefficient estimates for a nonlinear regression of the
responses in y on the predictors in X using the model
specified by fun.
Creating Graphical User
Interfaces (GUIs) using guide
Video: http://youtu.be/u06LwtHIh6Q
I hope this guide helps you get
started in MATLAB.
Feel free to talk to me or Arvind
about any issues you encounter
(but try online help first!)
