MATLAB(TM) Hypertext Reference, Copyright (c) 1995 Gerald Recktenwald, All rights reserved

Vectors


In MATLAB a vector is a matrix with either one row or one column. The distinction between row vectors and column vectors is essential. Many programming errors are caused by using a row vector where a column vector is required, and vice versa.

MATLAB vectors are used in many situations, e.g., creating x-y plots, that do not fall under the rubric of linear algebra. In these contexts a vector is just a convenient data structure. MATLAB still enforces the rules of linear algebra so paying attention to the details of vector creation and manipulation is always important.





Creating vectors

To create a vector you simply introduce it on the left hand side of an equal sign. Of course this means that the expression on the right side of the equal sign must evaluate to a vector. There are numerous ways to actually create a vector, each one having advantages in particular situations.
  1. using the the built-in functions ones, zeros, linspace, and logspace
  2. assigning a mathematical expressions involving vectors
  3. appending elements to a scalar
  4. using colon notation
Each method is useful and you will probably develop your own preferences. The first two methods will be described in this section. The last two methods are described in later sections.

Creating vectors with ones, zeros, linspace, and logspace

The ones, zeros linspace, and logspace functions allow for explicit creations of vectors of a specific size and with a prescribed spacing between the elements. These functions will be demonstrated by example without providing an exhaustive reference. Refer to the MATLAB manual (or help pages) for details.

To create a vector with one of these functions you must (atleast initially) decide how long do you want the vector to be. You must also decide whether the vector is a row or column vector.

The ones and zeros functions have two arguments. The first is the number of rows in the matrix you wish to create. The second is the number of columns. To create a row or a column vector set the appropriate argument of ones and zeros to one.

To create a row vector of length 5, filled with ones use 

	>> x = ones(1,5)
To create a column vector of length 5, filled with zeros use 
	>> y = zeros(5,1)

The linspace and logspace functions create vectors with linearly spaced or logarithmically spaced elements, respectively. Here are examples including the MATLAB output. 

	>> x = linspace(1,5,5)

	x =
	
	     1     2     3     4     5

	>> y = logspace(1,4,4)
	
	y =
	
	          10         100        1000       10000
The third argument of both linspace and logspace is optional. The third argument is the number of elements to use between the the range specified with the first and second arguments.

Assigning vector expressions to a vector

Once a vector has been created, it may be assigned to another vector. If the vector on the left of the equal sign does not exist it is created to fit the expression on the right hand side of the equal sign. 
	>> x = zeros(1,5);
	>> y = x;


Addressing vector elements

Individual elements of a vector can be addressed with a Fortran like subscript. For example 
	>> x = linspace(11,15,5);
	>> x(2)

	ans =
	
	     12
MATLAB automatically interprets the index as the appropriate row or column

Increasing the size of a vector (or scalar)

MATLAB allocates memory for all variables on the fly. This allows you to increase the size of a vector simply by assigning a value to an element that has not been previously used. 
	>> x = linspace(21,25,5)
	
	x =
	
	    21    22    23    24    25
	
	>> x(7) = -9
	
	x =
	
	    21    22    23    24    25     0    -9
This augmentation should be avoided for vectors involved in computations where speed is critical. Refer to Pre-allocating memory for vectors and matrices for a discussion of the performance issues.

Automatic augmentation of vectors does not allow you to refer to elements that have not yet been allocated. 

	>> y = linspace(21,25,5)
	
	y =
	
	    21    22    23    24    25
	
	>> y(7)
	???  Index exceeds matrix dimensions.


Colon notation

MATLAB colon notation is a compact way to refer to ranges of matrix elements. It is often used in copy operations and in the creation of vectors and matrices.

Colon notation can be used to create a vector as follows 

	>> x = xbegin:dx:xend
or 
	>> x2 = xbegin:xend
where xbegin and xend are the range of values covered by elements of the x vector, and dx is the (optional) increment. If dx is omitted a value of 1 (unit increment) is used. The numbers xbegin, dx, and xend need not be integers.

The preceding statements create row vectors. For example 

	>> x = 1:5

	x =
	
	1     2     3     4     5
To create a column vector, append the transpose operator to the end of the vector-creating expression 
	>> y = (1:5)'

	y =
	
	     1
	     2
	     3
	     4
	     5
Note that the colon expression needs to be enclosed in parentheses. Otherwise the transpose operator is applied to the value, 5, before the vector is created. (see Vector Transpose, below)

Using colon notation to create a vector requires you to specify the increment, whereas using the linspace command requires you to specify the total number of elements. The following commands show how to create the same vector with both approaches. 

	>> xbegin=1; xend=10;  nx=5;  dx=(xend-xbegin)/(nx-1);
	>> x1 = linspace(xbegin,xend,nx);
	>> x2 = xbegin:dx:xend;
The vectors are the same only if the increment dx corresponds to an integer number of elements. Prove this by repeating the preceding statments with nx=6;

When a colon expression appears in place of a vector (or matrix) index, the expression is a kind of implied do loop. The expression 

	istart:istop
refers to the range of numbers between istart and istop, inclusive. For example, the following statements create a row vector, x, and then copies the third through seventh elements of x into y. 
	>> x = linspace(31,40,10);
	>> y = x(3:7)

	y =
	
	    33    34    35    36    37
	
	>> y(3)
	
	ans =
	
	    35
The expression, y = x(3:7), copies the third through seventh elements of x into the first through fourth elements of y. If y did not already exist it is created by the assignment.



Vector operations

Mathematical operations involving vectors follow the rules of linear algebra. Addition and subtraction of vectors is performed with the + and - operators.

Not done yet!



[Preceding Section] [Master Outline] [Section Outline] [Next Section]