introduction to matlab session 2 simon o’keefe non-standard computation group [email protected]

45
Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group [email protected]

Post on 21-Dec-2015

227 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

Introduction to MATLAB

session 2Simon O’Keefe

Non-Standard Computation Group

[email protected]

Page 2: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

Content Writing scripts Flow control Writing and using functions Using cell arrays Creating structure arrays Plotting data

2

Page 3: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

1 Scripts

3

Page 4: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

1 Scripts Instead of typing each command into Matlab you can store

them in a file known as a script

To execute the commands in a script, type it’s name into the command prompt

Within the script, you have access to variables defined in the workspace.

Comments are denoted using the % symbol. Anything written after % is ignored by Matlab

mresult = mean(results) % calculate the mean of the data

4

Page 5: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

1 Scripts

5

To create a new script you can use the edit command>> edit

This can also be used to open an existing script>> edit script_name

Save the script, the filename will be the name of the script

To run the script type the name into the command prompt>> script_name

Page 6: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

2 Flow control

6

Page 7: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

2 Flow control For command

Use a for loop to repeat one or more statements

The end keyword tells Matlab where the loop finishes

You control the number of times a loop is repeated by defining the values taken by the index variable

This uses the colon operator again, so index values do not need to be integer

For example >> for i = 1:4

a(i) = i * 2 end

7

Page 8: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

2 Flow controlThe counter can be used to index different rows or columns E.g.

>> results = rand(10,3)

>> for i = 1:3

m(i) = mean(results(:, i))

end ..although you could do this in one step

m = mean(results);

8

Page 9: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

m(2) = mean(results(:, 2))m(3) = mean(results(:, 3))

i = 2i = 3

2 Flow control

>> for i = 1:3

m(i) = mean(results(:, i))

end

9

i = 1

m(1) = mean(results(:, 1))

Page 10: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

2 Flow control>> data = [4 14 6 11 3 14 8 17 17 12 10 18]>> cat = [1 3 2 1 2 2 3 1 3 2 3 1]

To work out the mean for each category you could type 3 commands:>> mdata(1) = mean(data(cat == 1))>> mdata(2) = mean(data(cat == 2))>> mdata(3) = mean(data(cat == 3))

Which is OK when there are a few categories but any more would create a lot of work

You can use a for loop instead>> for i = 1:3

mdata(i) = mean(data(cat == i)) end

The variable mdata will consist of 3 elements containing the mean of the values in data. The first element will contain the mean for category 1, second element the mean for category 2 and so forth.

10

Page 11: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

mdata(3) = mean(data(cat == 3))mdata(1) = mean(data(cat == 1))mdata(2) = mean(data(cat == 2))

2 Flow control>> for i = 1:3

mdata(i) = mean(data(cat == i))

end

11

i = 2i = 3i = 1

Page 12: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

2 Flow control The ‘if’ command is used with logical operators Again, the end command is used to tell Matlab where the

statement ends. For example, the following code loops through a matrix

performing calculations on each column >> for i = 1:size(results, 2)

m = results(:, i)

if m > 1

do something

else

do something different

end

end

12

Page 13: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

2 Flow control

13

Page 14: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

2 Flow control The ‘while’ command

>> while statement

commands

end

Waiting Reading

14

>> fid = fopen(‘results.txt'); fline = ‘’; while tline ~= -1 disp(tline) tline = fgetl(fid); end fclose(fid);

>> while mean(results) > 10 ind = results ~= max(results) results = results(ind) end

Page 15: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

2 Flow control We can use flow control to display results for each of

several experiments on separate plots.

15

mresult = mean(results)for i = 1:3 figure plot(results(:, i), ‘b.-’) hold on plot([1:10], repmat(mresult(i), 1, 10), ‘r-’) hold offend

Page 16: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

2 Flow control If we use flow control in a script we do not know the

size of the results matrix when it will be run. Instead, we make the script more general:

16

mresult = mean(results)for i = 1:size(results, 2) figure plot(results(:, i), ‘b.-’) hold on plot([1:size(results, 1)], repmat(mresult(i), 1, size(results,1)), ‘r-’) hold offend

Page 17: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

3 Functions

17

Page 18: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

3 Functions More functions:

pca – calculates the principle components of a set of data fft– performs the fast Fourier transform on a data set var – calculates the variance of the data repmat – replicates a matrix numel – returns the number of elements in a vector (or matrix) cumsum – calculates the cumulative summation sort – sorts a vector into ascending order floor & ceil – rounds data values down or up to the nearest whole

value

A list of the core functions that are available is located in Matlab’s help section. (Help Menu -> Matlab Help, in the right part of the window there is a Functions link)

18

Page 19: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

3 Functions>> sortrows(data, col)

19

Page 20: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

3 Functions Some functions can return two or more outputs.

If this is the case use the following command structure>> [output1, output2] = function_Name(input1)

For example>> rows = size(results)

Could be written:

>> [rows, cols] = size(results)

20

Page 21: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

3 Functions To find out what outputs a function can return use the

help command

21

Page 22: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

3 Functions>> [sdata, ind] = sortrows(data, col)

22

Page 23: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

3 Functions The function xlsread reads data from an Excel

spreadsheet>> [num, text] = xlsread(filename, sheet, ‘range’)

Only the filename parameter is required, the others are optional

The text output is optional, if it is not used only the data from the spreadsheet is loaded

23

Page 24: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

3 Functions

24

Page 25: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

3 Functions You can create your own function if you need to use

some calculation that is not provided by Matlab.

This done in a similar way to creating a script; use the same edit command to edit a function. >> edit function_Name

25

Page 26: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

3 Functions The flow of a function is the same as a script; commands

are carried out in the order that they appear and loops can be used.

However, a function does not have access to variables in the workspace. Instead, you pass the data to the function.

And once the function has finished it returns it’s results back to the output variable used when you called the function.

Any variables you create within the function are deleted.

26

Page 27: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

3 Functions When writing a function the first line must always be in

the form: function [outputvariable, outputvariable2] =

function_Name(inputvariable1, inputvariable2)

This line tells Matlab the name of the function and how it can be used

The function name must match the file name The output variable must be assigned a value inside the function. The input variables can be accessed inside the function using

their names.

27

Page 28: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

3 Functions The function:

function m = mymean(data)

m = sum(data) ./ size(data, 1)

Can be used in a script or at the command prompt using:>> cm = mymean(data)

28

Page 29: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

4 Cell Arrays

29

Page 30: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

4 Cell Arrays Standard arrays hold 1 value per element which is ideal for storing

results

However, if you try to store a string, each letter is treated as a vector element

For example >> str = ‘subject1’

[s, u, b, j, e, c, t, 1] >> str = [‘subject1’, ‘subject2’]

[s, u, b, j, e, c, t, 1, s, u, b, j, e, c, t, 2] >> str(1)

‘s’

Strings can not be stored and organised easily in vectors or matrices

Cell arrays allow you to do this

30

Page 31: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

4 Cell Arrays A cell array is similar to a normal array except that each cell

can contain a whole array or vector (or a single value)

And the item in each cell does not need to be the same size or even the same type

Curly brackets are used to define cell arrays All other operations work the same as with a normal array except

that you use curly brackets instead of round brackets

For example:>> labels = {‘string1’, ‘string2’, ‘string3’}>> labels{1}

‘string1’

31

Page 32: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

5 Plots

32

Page 33: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

5 Plots

Saving plots to a file>> print(fileformat, filename)

file format: ‘-dbitmap’ ‘-depsc’

You can also achieve this using the File -> Save menu in the figure’s window

The command line version is useful when you want to generate a lot of plots in a script which are saved automatically

33

Page 34: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

5 Plots

>> for i = 1:numel(results)

figure

plot(results(i).data, ‘b.-’)

hold on

plot([1:size(results(i).data)], repmat(results(i).mdata), ‘r-’)

hold off

print(‘-bmp’, [‘c:\users\tom\desktop\’, results(i).label])

end

34

Page 35: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

5 Plots

Subplots Multiple plots can be placed within the same window This is achieved using the subplot command The window is split into a grid the size of which is specified

when entering the command

>> subplot(2,2,1)

This creates a grid 2 x 2 in size (4 plots) and sets the current plot to the first of these.

35

Page 36: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

5 Plots

>> subplot(2,2,1)

36

43

21

Page 37: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

5 Plots

>> figure

cols = 4

rows = ceil(numel(results) / cols)

for i = 1:numel(results)

subplot(rows, cols, i)

plot(results(i).data, ‘b.-’)

hold on

plot([1:size(results(i).data)], repmat(results(i).mdata), ‘r-’)

hold off

end

37

Page 38: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

5 Plots

38

Page 39: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

5 Plots Handles allow you to create a plot and then edit the properties later.

A handle is created when you create a plot or an object within the plot (for example title, legend) >> h = plot(data)This returns a handle to the line/s plotted, now you can change the line

style, colour, width etc.

The handle is stored in a variable which can then be used to edit the properties with the set command

For example, the position of a plot’s legend can be changed using handles >> plot(data) >> h = legend('line') >> set(h,'Location','South')

39

Page 40: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

5 Plots Inside/outside Best/Bestoutside

40

Northwest NortheastNorth

South SoutheastSouthwest

EastWest

Page 41: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

5 Plots

Colour bars can be repositioned in the same way:

>> h = colorbar

>> set(h, ‘Location’, ‘North’)

41

Page 42: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

5 Plots The figure function also returns a handle, you can

use the set function to change the figure title: >> h = figure >> set(h, 'Name', 'Subject1')

42

Page 43: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

5 Plots

Surface plot

>> surf(matrix)

>> colorbar

>> axis([1 50 1 50 0 1])

>> title('Surface Plot');

>> xlabel('x')

>> ylabel('y')

>> zlabel('z')

43

Page 44: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

5 Plots

Contour

>> contour(data)

>> colorbar

>> xlabel('x')

>> ylabel('y')

>> zlabel('z')

>> title('Contour Plot');

>> axis([1 50 1 50])

44

Page 45: Introduction to MATLAB session 2 Simon O’Keefe Non-Standard Computation Group sok@cs.york.ac.uk

5 Plots All of these can be use in conjunction with subplot,

therefore, you can display several different plot types in one figure

45