Brief Octave Cheat Sheet for Coursera Machine Learning Course by Stanford University
I use R with Python a lot, Octave is the chosen language in Coursera course: Machine Learning by Stanford University.
So this article will only cover necessary concept to finish this Machine Learning course.
Index
In Octave, matrix and vector are indexed from 1, which differs from many other languages.
Output
One line not ending with a semicolon will print the result to output, with semicolon with suppress that output.
Or use disp(i); or sprintf(i)
String
Compare
strcmp, strmatch
Cell-array
Acess
ca = cell(2,1); % create cell array
ca{1} = 'abc'; % assign to first element
ca{2} = 'def'; % assign to second element
ca{1}; % access first element
Range
1:10 will create a 1x10 matrix or say 10-element vector with numbers from 1 to 10. 1:2:10 will create vector with each other numbers from 1 to 10, i.e. [1 3 5 7 9]. The middle 2 is the specified step.
Matrix, Vector
Assignment
% assign a matrix to A:
A = [1 2; 3 4; 5 6]
A =
1 2
3 4
5 6
% assign to second column
>> A(:,2) = [10 11 12]
% space, comma or semicolon doesn't matter here
>> A(:,2) = [10, 11, 12]
>> A(:,2) = [10; 11; 12]
Access
% access A at first row and second column.
>> A(1,2)
ans = 2
% access second row, here colon refers to all columns
>> A(2,:)
ans =
3 4
% access second column
>> A(:,2)
ans =
2
4
6
a(2) # result is a scalar
a(1:2) # result is a row vector
a([1; 2]) # result is a column vector
a = [1, 2; 3, 4]
all of the following expressions are equivalent and select the first row of the matrix.
a(1, [1, 2]) # row 1, columns 1 and 2
a(1, 1:2) # row 1, columns in range 1-2
a(1, :) # row 1, all columns
a(1:end/2) # first half of a => [1, 2]
a(end + 1) = 5; # append element
a(end) = []; # delete element
a(1:2:end) # odd elements of a => [1, 3]
a(2:2:end) # even elements of a => [2, 4]
a(end:-1:1) # reversal of a => [4, 3, 2 , 1]
Fill
>> A = ones(2, 3) # ones or zeros
A =
1 1 1
1 1 1
>> rand(2, 3)
ans =
0.47210 0.10022 0.35182
0.69316 0.71345 0.71179
Concatenate
>> B = [20 21; 22 23; 24 25]
>> C = [A B] % concatenate A B horizontally
ans =
1 2 20 21
3 4 22 23
5 6 24 25
>> D = [A; B] % concatenate A B vertically
ans =
1 2
3 4
5 6
20 21
22 23
24 25
Transpose
>> A' % transpose
ans =
1 3 5
2 4 6
Max, Min
>> max(magic(4)) % return every column's max
ans =
16 14 15 13
>> max(magic(4), [], 2) % return every row's max
ans =
16
11
12
15
>> [val, ind] = max(magic(4)) % retuns every column's max and their index in corresponding column
val =
16 14 15 13
ind =
1 4 4 1
>> max(max(A)); % max element in matrix
>> A(:) % convert matrix into one column
ans =
1
3
5
2
4
6
>> max(A(:)) % max element in matrix
ans =
6
Sum
>> sum(A) % sum of column
>> sum(A, 2) % sum of row
Sum of diagonals in a square matrix
>> M = magic(4);
>> sum(sum(M.*eye(4))) % sum of diagonal top left to bottom right
ans = 34
>> sum(sum(M.* flipud(eye(4) ))) % sum of diagonal bottom left to top right
ans = 34
Flip
>> flipud(A) $ flip matrix upside down
ans =
5 6
3 4
1 2
Matrix select
A(A==2)
Reshape
>> reshape(A, 2, 3)
ans =
1 5 4
3 2 6
Functions & control statements
Functions are saved in files with the file-ending .m for MATLAB.
function y = function_name(x1, ...
x2) % x2 is optional
if ~exist('x2', 'var') || isempty(x2)
x2 = 1;
end
y = x1 + x2;
% y is the return value
% x1 is a parameter
% is also possible to return multiple values
function [y1, y2] = function_name(x1)
y1 = x1^2
y2 = x1^3
>> for i=1:10
>> disp(i)
>> end;
>> i = 1;
>> while (i ~= 10)
>> disp(i);
>> i = i+1;
>> endwhile;
% i = 10
>> if (i == 10)
>> sprintf('yes')
>> else
>> sprintf('no')
>> endif
ans = yes
Anonymous function
@(x1, x2) another_func(x1, x2) % anonymous function, just like Python lambda
Logic operations
not equal ~=
logical AND &&
logical OR ||
logical XOR xor(1,0)
Reference:
https://gist.github.com/obstschale/7320846
http://folk.ntnu.no/joern/itgk/refcard-a4.pdf