Introduction to Matlab
Control Structures

• Most of the basic concepts we learned in C/C++ apply in Matlab. Only some of the implementation details differ.
• Relational Operators - >,<,>=,<=,= =, ~ =
  • Note that "not equals" uses a tilde, ~, in place of the ! in C++
  • Applied to matrices produces matrices of zeros and ones
    • These zeros and ones can then be used as indices to refer to positions within matrices.
    • Example:
      • m = magic( 5 ); % Five by five magic square, values 1 to 25
      • big = ( m >= 20 ); % Ones indicate positions where m is >= 20
      • m( big ) = 0; % Changes values of only those positoins to zero.
      • m( 5 < m & m < 10 ) = 42.0; % Without intermediary array. Note single &, not double &&.
      • twoTone = 255 * big; % Array of all 255 or zero, based on whether m >= 20 ( originally ).
• Logical Operators - &, |, ~ for AND, OR, NOT
  • Single &, | does elementwise comparison, continuing for all elements.
  • Double &&, || does short-circuiting, similar to C++, only works on scalars.
• if - elseif - else - end
  • No curly braces, but the "end" command is required.
  • elseif and else are optional.
  • No parentheses required, but they don't hurt either.
• while loops
  • Condition controlled loops, just as in C/C++
  • No curly braces, but a closing "end" is required.
  • No parentheses required, but they don't hurt either.
  • demonstrated in epsilonExample.m
  • Matlab does not have a do-while loop.
• for loops
  • NOT a counter-controled loop as in C, though that behaviour can be mimiced.
  • The for loop in Matlab is a list-controlled loop
  • Syntax:

  for index = list

  statements

  end

  • index takes on one value from the list for each iteration.
  • If the "list" is a 2-D matrix,then index is a column vector inside the loop.
• for loops and while loops support break and continue
• Matlab does not have the conditional operator, ?: . ( Also no ++, --, +=, etc. )
• find - Generates indices of non-zero elements.
  • Example: find( x > 0 ) returns indices for which x is positive.
  • Normally treats 2-D matrices as one long column
  • [ rows, columns ] = find( ... ) returns separate row and column information.
  • ( However results of relational operators can be used directly, as shown above. )
• Matlab has a switch construct, not covered in our current book.
  • Syntax:

                      switch switch_expression
                      		case case_expression
                                    statements
                            case case_expression
                                    statements
                               :
                            otherwise
                                 statements
                     end

  • Automatically breaks at the end of each case.
  • Cases can be combined by using a list for the case expression
  • No braces, but a closing "end" is required
  • No colon after the case expression
  • Parentheses not required around the switch expression, but they don't hurt either
  • Matlab uses "otherwise" instead of "default"
• menus also not covered in our current book - combined with an infinite loop and switch in menuExample.m