ChBE 303

Introduction to Matlab
Part 1
Robert A Pinnick
Fall 2006
 Outline
 Essentials
 Philosophy of Data Storage
 Mathematical Operators
 Elementary
 1-Dimensional Arrays ( Vectors )
 2-Dimensional Arrays ( Matrix )
 Structure Operations
 Cell Operations
 Example Code
 Linear Regression
 Numerical Integration
 Numerical Solvers
 Statistics
 Essentials
 Syntax & Logic Help
 >> help <function_name>
 help fzero
 >> lookfor <keyword>
 lookfor bessel
 www.mathworks.com
 Support (Documentation) & Forum Sections
 www.google.com
 Useful Tips
 >> clear (variable)
 Clears All Memory or Specified Variable
 >> clc
 Clears Command Window Screen
 Philosophy of Data Storage
 Numerical Values (Ex. 5, 3.14159, 2+i)
 Integers
 Floating (Single & Double)
 Complex
 Character Strings (Ex. ‘asdf jkl;’)
 Structures
 Cells
 Boolean ( True / False )
 Elementary Operators
 Assignment (=)  Multiplication (*)
 >> X = 1;  >> Z = 2 * Y
 >> Y = 3.14159;  Z = 6.28318
 Addition (+)  Division (/)
 >> Z = X + Y  >> Z = 1 / 4
 Z = 4.14159  Z = 0.25
 Subtraction (-)  Power (^)
 >> Z = X – Y  >> Z = 5 ^ 3
 Z = -2.14159  Z = 125
 Vector Operators
– Scalar Operations –
 Assignment  Index
 Column  >> Z = X(3)
 >> X = [ 1 ; 2 ; 3 ];  Z=3
 Row  >> Z = X(4)
 >> Y = [ 1 , 2 , 3 ]; –OR–
 >> Y = [ 1 2 3 ];
>> Z = X(0)
 Unique Commands  ERROR !!
 linspace( Initial , Final , # Points )
 >> Z = linspace( 5 , 20 , 4 )
 >> Z = X(1:2)
 Z = [ 5 , 10 , 15 , 20 ];  Z=[1;2]
 logspace( Initial , Final , # Points )  >> Z = X([ 1 3 ])
 Initial:Step:Final  Z=[1;3]
 >> Z = 1:-0.25:0
 Z = [ 1 , 0.75 , 0.5 , 0.25 , 0 ]
 Vector Operators
– Scalar Operations –
 Addition (+)  Summation
 >> Z = X + 2 n

 Z=[3;4;5] x
i 1
i

 Subtraction (-)
 >> X = [ 1 , 2 , 3 , 4 ];
 Multiplication (*) >> Z = sum( X )
 >> Z = 2 * X  Z = 10
 Z=[2;4;6]
 Product
 Division (/) n
 >> Z = X / 2
 Z = [ 0.5 ; 1 ; 1.5 ]
x
i 1
i

 >> Y = [ 1 ; 2 ; 3 ; 4 ];
>> Z = prod( Y )
 Z = 24
 Vector Operators
 Inner Product
 y1 
 
 Euclidean Norm
    n
x  y  x1  xi  xn  yi    xi yi n
  
  i 1
  x  x  x   xi2
 yn  i 1

>> X = [ 1 , 2 , 3 ];
>> X = [ 1 , 2 , 3 ];


>> Z = X * X’
>> Y = [ 1 ; 2 ; 3 ];  Z = 14

>> Z = X * Y
Z = 14
Cross Product


 Transpose  
x  y  x2 y3  y2 x3 x3 y1  y3 x1 x1 y2  y1 x2 
 x1 
 
 
 >> X = [ 1 , 2 , 3 ];
x1  xi  xn 
T
  xi  >> Y = [ 3 , 2 , 1 ];
 
  >> Z = cross( X , Y )
 xn 
 Z=
 >> Z = X’
 Z=[1;2;3]
 Vector Operators
– Element Wise –
 Addition (+)  Multiplication (.*)
 >> X = [ 1 , 2 , 3 ];  >> X = [ 1 , 2 , 3 ];
>> Y = [ 2 , 4 , 6 ]; >> Y = [ 2 , 4 , 6 ];
>> Z = X + Y >> Z = X.*Y
 Z = [ 3 , 6 , 9 ];  Z = [ 2 , 8 , 18 ];
 Subtraction (-)  Division (./)
 Power (.^)
 >> Z = X.^2
 Z = [ 1 , 4 , 9 ];
 Numerical Integration
Numerical Integration
 Data ( y = x^2 )
40
 ( -6, 36 ) 35
30
 ( -3 , 9 ) 25
20

Y
 (0,0) 15
10
 (3,9) 5
0
 ( 6 , 36 ) -6 -4 -2 0 2 4 6
X
 Code
 >> X = [ -6 , -3 , 0 , 3 , 6 ];
>> Y = [ 36 , 9 , 0 , 9 , 36 ];  Algorithms
>> Z = sum( trapz( X , Y ) );  Trapezoidal Rule
 Z = 162  Simpson’s Rule
 >> Z = QUAD( INLINE( ‘x.^2‘ ) , -6 …  Etc.
6)
 Z = 144
 Statistics Functions
 Mean n  Standard Deviation

x
>> X = rand(10); x  i 1 i
n
 >> Y = mean( X ) n 
 ix  x 2

 i 1
 Median n
 >> Y = mean( X )  >> Y = std( X )
 Maximum  Sort
 >> Y = max( X )  >> Y = sort( X )
 Minimum  Find
 >> Y = min( X )  >> Y = X( find( X > 0.5 ) )
 Matrix Operators
– Scalar Operations –
 Assignment  Unique Functions
 >> X = [ 1 , 2 ; 3 , 4 ];  Identity Matrix
 Concatenation eye( Column , Row )
 >> X = 1:3;  >> X = eye(2,2)
>> Y = [ X ; 2*X ]
 Y=[1,2,3;2,4,6]
 X=[1,0;0,1]

 Index  Zeros & Ones Matrix

 zeros( Column , Row )
 >> Z = X(1,2)
–OR–  ones( Column , Row )
>> Z = X(3)  Random # ( 0 – 1 )
 Z=3 rand( Column , Row )
 >> Z = X(2,3)  >> X = rand( 2 , 2 )
–OR–  X = [ 0.5234 , 0.9246 ;
>> Z = X(5) 0.2862 , 0.7378 ]
 ERROR !!
 Matrix Operators
– Scalar / Element Wise –
 Scalar Operations  Element Wise Operations
 Addition (+)  Addition (+)
 >> Z = X + 2  >> X = [ 1 , 2 ; 3 , 4 ];
 Z=[3,4;5,6] >> Y = [ 4 , 3 ; 2 , 1 ];
>> Z = X+Y
 Subtraction (-)  Z=[5,5;5,5]
 Multiplication (*)  Subtraction (-)
 >> Z = 2 * X  Multiplication
 Z=[2,4;6,8]  >> Z = X.*Y
 Division (/)  Z=[4,6;6,4]
 Division
 Matrix Operations
– Linear Algebra … A x = b –
 b=?  x=?
 >> A = [ 1 , 2 ; 3 , 4 ];  >> A = [ 1 , 2 ; 3 , 4 ];
>> x = [ 1 ; 2 ];
>> b = A*x >> b = [ 5 ; 11 ];
 b = [ 5 ; 11 ]; >> x = A\b
 x = [ 1 ; 2 ];
 n
  A x

  b1 
 Algorithms …
 A1,1  A1,i  A1,n   x1  1, i i
 
Gaussian Elimination
i 1
               
   

n

  
A  x   A i,1  A i,i  A i, n  xi    A j,i xi   bi   b


    
  
      i 1     
    LU Factorization
A n,1  A n,i  A n,n   xn   n  b 
  A n,i xi   n 
 i 1 
 Etc.
 Matrix Operations
 Inverse  Eigenstates
 >> X = [ 1 2 3 ; 2 3 1 ; 3 1
2]; Ax  x
 >> Y = inv(X);
 >> Z = X*Y  >> [ V , D ] = eig(Z)
 Z=[100;010;001]
 Eigenvalues …
 D = [-1.73, 1.73, 6.00]
 Determinant  Eigenvectors …
 >> Z = det( X ) 
V = [ 0.79 0.21 0.58 ; -0.21
 Z = -18 -0.79 0.58 ; -0.58 0.58
 EXAMPLE (2 x 2) 0.58]
X  x1,1 x2, 2  x2,1 x1, 2
 EXAMPLE (3 x 3)
X  x1,1 x2, 2 x3,3  x1,1 x3, 2 x2,3  x1, 2 x2,3 x3,1  
x1, 2 x2,1 x3,3  x1,3 x3, 2 x2,1  x1,3 x2, 2 x2,1
 Linear Regression
 Data …  QR Factorization
 ( 2, 3 )  >> X = [ 2 ; 4 ; 5 ; 6 ; 7 ] ;
>> Y = [ 3 ; 7 ; 10 ; 10 ; 12 ];
 ( 4, 7 ) >> A = [ X , ones(size(X)) ];
 ( 5, 10 ) >> COEF = A \ Y
 COEF = [ 1.7838 , -0.1622 ]
 ( 6, 10 )
 ( 7, 12 )  Polyfit
 >> COEF = polyfit( X , Y , 1 )
 COEF = [ 1.7838 , -0.1622 ]
 Concept of Transform
 Instead of Nonlinear Regression …
 Linearize the Equation

 Exponentials
 Y = a*e^(b*X)
 ln(Y) = ln(a) + b*X

 Power Law
 Y = a * X^b …
 ln(Y) = ln(a) + b*ln(X)
 Power Law Regression
– Example –
 Regress Data  Calling Code
( y = 2*x^2 + noise )  >> x = [ 1 , 2 , 3 , 4 , 5 , 6 ];
 (1,3) >> y = [ 3 , 8 , 17 , 33 , 50 , 71 ];
 (2,8) >> COEF = powerfit(x,y)
 ( 3 , 17 )  COEF = [ 1.7971 , 2.6552 ]
 ( 4 , 33 )
Power Law Regression
 ( 5 , 50 )
100
 ( 6 , 72 )
 POWERFIT
function COEF = powerfit( x , y )
10
y

ln_x = log( x );
ln_y = log( y );
COEF = polyfit( x , y , 1 );
COEF(2) = exp( COEF(2) );
1
1 10
x
 Character Strings
 Assignment  Comparison
 >> X = ‘HELLO’;  >> X = ‘HI’
 >> Y = 2.7183;  >> Y = ‘HE’
 >> Z = num2str( Y )  >> Z = strcmp( X, Y )
 Z = ‘2.7183’  Z=0
 Concatenation
 >> X = strcat( ‘HELLO’ , ‘ WORLD’ )
-OR-
>> X = [‘HELLO’ ‘WORLD’]
 X = ‘HELLOWORLD’
 Structures

 Assignment ( Variable.Property = XYZ )

 >> Person(1).Name = ‘BOB’
 >> Person(1).Addres = ’99 Sunset Blvd’
 >> Person(1).Phone = 3334444
 >> Person(2).Name = ‘JANE’
 >> Person(2).Address = ‘1750 Kirby Dr’
 … etc.
 Cells
 Assignment
 >> X = { 1:4 , ‘HI’ ; [ 1 , 0 ; 0 , 1 ] , 3.1459 };
 Index
 >> Z = X{1,1}
-OR-
>> Z = X{1}
 Z=[1234]
 >> Z = X{1,2}(1)
 Z = ‘H’
 >> Z = X{2,1}(1,2)
 Z=0
 Nonlinear Root Finding
 FZERO  Algorithms
( exp( 1 / x ) sin( x ) / x^2 )  Bisection Method
 >> FUN = inline(  Newton’s Method
‘exp( 1 / x ) sin( x ) / x^2’ );  Etc.
 >> X0 = 3;
y=ex sin(x)x -2
150
 >> X = fzero( FUN , X0 )

 X = 3.1416 100

50

 ROOTS – Polynomials 0
Y

( Y = -2X^3 + 4.7*X + 0.25 )

-50
 >> X = [ -2 , 0 , 4.7 , 0.25 ];

 >> Z = roots( X ); -100

 Z = [1.56 -1.51 -0.05]

-150
1 2 3 4 5 6 7 8 9 10
X