I am trying to write code to get the 'N-dimensional product' of vectors. So for example, if I have 2 vectors of length L, x & y, then the '2-dimensional product' is simply the regular vector product, R=x*y', so that each entry of R, R(i,j) is the product of the i'th element of x and the j'th element of y, aka R(i,j)=x(i)*y(j).
The problem is how to elegantly generalize this in matlab for arbitrary dimensions. This is I had 3 vectors, x,y,z, I want the 3 dimensional array, R, such that R(i,j,k)=x(i)*y(j)*z(k).
Same thing for 4 vectors, x1,x2,x3,x4: R(i1,i2,i3,i4)=x1(i1)*x2(i2)*x3(i3)*x4(i4), etc...
Also, I do NOT know the number of dimensions beforehand. The code must be able to handle an arbitrary number of input vectors, and the number of input vectors corresponds to the dimensionality of the final answer.
Is there any easy matlab trick to do this and avoid going through each element of R specifically?
Thanks!
I think by "regular vector product" you mean outer product.
In any case, you can use the ndgrid function. I like this more than using bsxfun as it's a little more straightforward.
% make some vectors
w = 1:10;
x = w+1;
y = x+1;
z = y+1;
vecs = {w,x,y,z};
nvecs = length(vecs);
[grids{1:nvecs}] = ndgrid(vecs{:});
R = grids{1};
for i=2:nvecs
R = R .* grids{i};
end;
% Check results
for i=1:10
for j=1:10
for k=1:10
for l=1:10
V(i,j,k,l) = R(i,j,k,l) == w(i)*x(j)*y(k)*z(l);
end;
end;
end;
end;
all(V(:))
ans = 1
The built-in function bsxfun is a fast utility that should be able to help. It is designed to perform 2 input functions on a per-element basis for two inputs with mismatching dimensions. Singletons dimensions are expanded, and non-singleton dimensions need to match. (It sounds confusing, but once grok'd it useful in many ways.)
As I understand your problem, you can adjust the dimension shape of each vector to define the dimension that it should be defined across. Then use nested bsxfun calls to perform the multiplication.
Example code follows:
%Some inputs, N-by-1 vectors
x = [1; 3; 9];
y = [1; 2; 4];
z = [1; 5];
%The computation you describe, using nested BSXFUN calls
bsxfun(#times, bsxfun(#times, ... %Nested BSX fun calls, 1 per dimension
x, ... % First argument, in dimension 1
permute(y,2:-1:1) ) , ... % Second argument, permuited to dimension 2
permute(z,3:-1:1) ) % Third argument, permuted to dimension 3
%Result
% ans(:,:,1) =
% 1 2 4
% 3 6 12
% 9 18 36
% ans(:,:,2) =
% 5 10 20
% 15 30 60
% 45 90 180
To handle an arbitrary number of dimensions, this can be expanded using a recursive or loop construct. The loop would look something like this:
allInputs = {[1; 3; 9], [1; 2; 4], [1; 5]};
accumulatedResult = allInputs {1};
for ix = 2:length(allInputs)
accumulatedResult = bsxfun(#times, ...
accumulatedResult, ...
permute(allInputs{ix},ix:-1:1));
end
Related
Suppose I have a matrix A of dimension Nx(N-1) in MATLAB, e.g.
N=5;
A=[1 2 3 4;
5 6 7 8;
9 10 11 12;
13 14 15 16;
17 18 19 20 ];
I want to transform A into an NxN matrix B, just by adding a zero diagonal, i.e.,
B=[ 0 1 2 3 4;
5 0 6 7 8;
9 10 0 11 12;
13 14 15 0 16;
17 18 19 20 0];
This code does what I want:
B_temp = zeros(N,N);
B_temp(1,:) = [0 A(1,:)];
B_temp(N,:) = [A(N,:) 0];
for j=2:N-1
B_temp(j,:)= [A(j,1:j-1) 0 A(j,j:end)];
end
B = B_temp;
Could you suggest an efficient way to vectorise it?
You can do this with upper and lower triangular parts of the matrix (triu and tril).
Then it's a 1 line solution:
B = [tril(A,-1) zeros(N, 1)] + [zeros(N,1) triu(A)];
Edit: benchmark
This is a comparison of the loop method, the 2 methods in Sardar's answer, and my method above.
Benchmark code, using timeit for timing and directly lifting code from question and answers:
function benchie()
N = 1e4; A = rand(N,N-1); % Initialise large matrix
% Set up anonymous functions for input to timeit
s1 = #() sardar1(A,N); s2 = #() sardar2(A,N);
w = #() wolfie(A,N); u = #() user3285148(A,N);
% timings
timeit(s1), timeit(s2), timeit(w), timeit(u)
end
function sardar1(A, N) % using eye as an indexing matrix
B=double(~eye(N)); B(find(B))=A.'; B=B.';
end
function sardar2(A,N) % similar to sardar1, but avoiding slow operations
B=1-eye(N); B(logical(B))=A.'; B=B.';
end
function wolfie(A,N) % using triangular parts of the matrix
B = [tril(A,-1) zeros(N, 1)] + [zeros(N,1) triu(A)];
end
function user3285148(A, N) % original looping method
B = zeros(N,N); B(1,:) = [0 A(1,:)]; B(N,:) = [A(N,:) 0];
for j=2:N-1; B(j,:)= [A(j,1:j-1) 0 A(j,j:end)]; end
end
Results:
Sardar method 1: 2.83 secs
Sardar method 2: 1.82 secs
My method: 1.45 secs
Looping method: 3.80 secs (!)
Conclusions:
Your desire to vectorise this was well founded, looping is way slower than other methods.
Avoiding data conversions and find for large matrices is important, saving ~35% processing time between Sardar's methods.
By avoiding indexing all together you can save a further 20% processing time.
Generate a matrix with zeros at diagonal and ones at non-diagonal indices. Replace the non-diagonal elements with the transpose of A (since MATLAB is column major). Transpose again to get the correct order.
B = double(~eye(N)); %Converting to double since we want to replace with double entries
B(find(B)) = A.'; %Replacing the entries
B = B.'; %Transposing again to get the matrix in the correct order
Edit:
As suggested by Wolfie for the same algorithm, you can get rid of conversion to double and the use of find with:
B = 1-eye(N);
B(logical(B)) = A.';
B = B.';
If you want to insert any vector on a diagonal of a matrix, one can use plain indexing. The following snippet gives you the indices of the desired diagonal, given the size of the square matrix n (matrix is n by n), and the number of the diagonal k, where k=0 corresponds to the main diagonal, positive numbers of k to upper diagonals and negative numbers of k to lower diagonals. ixd finally gives you the 2D indices.
function [idx] = diagidx(n,k)
% n size of square matrix
% k number of diagonal
if k==0 % identity
idx = [(1:n).' (1:n).']; % [row col]
elseif k>0 % Upper diagonal
idx = [(1:n-k).' (1+k:n).'];
elseif k<0 % lower diagonal
idx = [(1+abs(k):n).' (1:n-abs(k)).'];
end
end
Usage:
n=10;
k=3;
A = rand(n);
idx = diagidx(n,k);
A(idx) = 1:(n-k);
As always trying to learn more from you, I was hoping I could receive some help with the following code.
I need to accomplish the following:
1) I have a vector:
x = [1 2 3 4 5 6 7 8 9 10 11 12]
2) and a matrix:
A =[11 14 1
5 8 18
10 8 19
13 20 16]
I need to be able to multiply each value from x with every value of A, this means:
new_matrix = [1* A
2* A
3* A
...
12* A]
This will give me this new_matrix of size (12*m x n) assuming A (mxn). And in this case (12*4x3)
How can I do this using bsxfun from matlab? and, would this method be faster than a for-loop?
Regarding my for-loop, I need some help here as well... I am not able to storage each "new_matrix" as the loop runs :(
for i=x
new_matrix = A.*x(i)
end
Thanks in advance!!
EDIT: After the solutions where given
First solution
clear all
clc
x=1:0.1:50;
A = rand(1000,1000);
tic
val = bsxfun(#times,A,permute(x,[3 1 2]));
out = reshape(permute(val,[1 3 2]),size(val,1)*size(val,3),[]);
toc
Output:
Elapsed time is 7.597939 seconds.
Second solution
clear all
clc
x=1:0.1:50;
A = rand(1000,1000);
tic
Ps = kron(x.',A);
toc
Output:
Elapsed time is 48.445417 seconds.
Send x to the third dimension, so that singleton expansion would come into effect when bsxfun is used for multiplication with A, extending the product result to the third dimension. Then, perform the bsxfun multiplication -
val = bsxfun(#times,A,permute(x,[3 1 2]))
Now, val is a 3D matrix and the desired output is expected to be a 2D matrix concatenated along the columns through the third dimension. This is achieved below -
out = reshape(permute(val,[1 3 2]),size(val,1)*size(val,3),[])
Hope that made sense! Spread the bsxfun word around! woo!! :)
The kron function does exactly that:
kron(x.',A)
Here is my benchmark of the methods mentioned so far, along with a few additions of my own:
function [t,v] = testMatMult()
% data
%{
x = [1 2 3 4 5 6 7 8 9 10 11 12];
A = [11 14 1; 5 8 18; 10 8 19; 13 20 16];
%}
x = 1:50;
A = randi(100, [1000,1000]);
% functions to test
fcns = {
#() func1_repmat(A,x)
#() func2_bsxfun_3rd_dim(A,x)
#() func2_forloop_3rd_dim(A,x)
#() func3_kron(A,x)
#() func4_forloop_matrix(A,x)
#() func5_forloop_cell(A,x)
#() func6_arrayfun(A,x)
};
% timeit
t = cellfun(#timeit, fcns, 'UniformOutput',true);
% check results
v = cellfun(#feval, fcns, 'UniformOutput',false);
isequal(v{:})
%for i=2:numel(v), assert(norm(v{1}-v{2}) < 1e-9), end
end
% Amro
function B = func1_repmat(A,x)
B = repmat(x, size(A,1), 1);
B = bsxfun(#times, B(:), repmat(A,numel(x),1));
end
% Divakar
function B = func2_bsxfun_3rd_dim(A,x)
B = bsxfun(#times, A, permute(x, [3 1 2]));
B = reshape(permute(B, [1 3 2]), [], size(A,2));
end
% Vissenbot
function B = func2_forloop_3rd_dim(A,x)
B = zeros([size(A) numel(x)], 'like',A);
for i=1:numel(x)
B(:,:,i) = x(i) .* A;
end
B = reshape(permute(B, [1 3 2]), [], size(A,2));
end
% Luis Mendo
function B = func3_kron(A,x)
B = kron(x(:), A);
end
% SergioHaram & TheMinion
function B = func4_forloop_matrix(A,x)
[m,n] = size(A);
p = numel(x);
B = zeros(m*p,n, 'like',A);
for i=1:numel(x)
B((i-1)*m+1:i*m,:) = x(i) .* A;
end
end
% Amro
function B = func5_forloop_cell(A,x)
B = cell(numel(x),1);
for i=1:numel(x)
B{i} = x(i) .* A;
end
B = cell2mat(B);
%B = vertcat(B{:});
end
% Amro
function B = func6_arrayfun(A,x)
B = cell2mat(arrayfun(#(xx) xx.*A, x(:), 'UniformOutput',false));
end
The results on my machine:
>> t
t =
0.1650 %# repmat (Amro)
0.2915 %# bsxfun in the 3rd dimension (Divakar)
0.4200 %# for-loop in the 3rd dim (Vissenbot)
0.1284 %# kron (Luis Mendo)
0.2997 %# for-loop with indexing (SergioHaram & TheMinion)
0.5160 %# for-loop with cell array (Amro)
0.4854 %# arrayfun (Amro)
(Those timings can slightly change between different runs, but this should give us an idea how the methods compare)
Note that some of these methods are going to cause out-of-memory errors for larger inputs (for example my solution based on repmat can easily run out of memory). Others will get significantly slower for larger sizes but won't error due to exhausted memory (the kron solution for instance).
I think that the bsxfun method func2_bsxfun_3rd_dim or the straightforward for-loop func4_forloop_matrix (thanks to MATLAB JIT) are the best solutions in this case.
Of course you can change the above benchmark parameters (size of x and A) and draw your own conclusions :)
Just to add an alternative, you maybe can use cellfun to achieve what you want. Here's an example (slightly modified from yours):
x = randi(2, 5, 3)-1;
a = randi(3,3);
%// bsxfun 3D (As implemented in the accepted solution)
val = bsxfun(#and, a, permute(x', [3 1 2])); %//'
out = reshape(permute(val,[1 3 2]),size(val,1)*size(val,3),[]);
%// cellfun (My solution)
val2 = cellfun(#(z) bsxfun(#and, a, z), num2cell(x, 2), 'UniformOutput', false);
out2 = cell2mat(val2); % or use cat(3, val2{:}) to get a 3D matrix equivalent to val and then permute/reshape like for out
%// compare
disp(nnz(out ~= out2));
Both give the same exact result.
For more infos and tricks using cellfun, see: http://matlabgeeks.com/tips-tutorials/computation-using-cellfun/
And also this: https://stackoverflow.com/a/1746422/1121352
If your vector x is of lenght = 12 and your matrix of size 3x4, I don't think that using one or the other would change much in term of time. If you are working with higher size matrix and vector, now that might become an issue.
So first of all, we want to multiply a vector with a matrix. In the for-loop method, that would give something like that :
s = size(A);
new_matrix(s(1),s(2),numel(x)) = zeros; %This is for pre-allocating. If you have a big vector or matrix, this will help a lot time efficiently.
for i = 1:numel(x)
new_matrix(:,:,i)= A.*x(i)
end
This will give you 3D matrix, with each 3rd dimension being a result of your multiplication. If this is not what you are looking for, I'll be adding another solution which might be more time efficient with bigger matrixes and vectors.
I have a matrix 2NxN.
And I want get some parametrs by this matrix. For example it:
How, I can do it?
You may want to break your 12x6 matrix, into two 6x6 matrix; let's say: Z and Zb (last one for z bar). Odd rows are Z and evens are Zb.
Considering M to be the combined matrices:
Z = M(1:2:end,:)
Zb = M(2:2:end,:)
read about the colon(:) operator and end to see what 1:2:end means.
Hope it helps.
From what I understand here are the first three:
% Random Matrix
% Needs to be defined before the functions since the functions look for
% the m variable
m = rand(12,6);
% Function 1
p = #(i,j) sign(m(i,j)+m(i+1,j)) * max(abs(m(i,j)),abs(m(i+1,j)));
p(2,2)
% Function 2 - Avg of row
pavg = #(i) mean(m(i,:));
pavg(2)
% Function 3
c = #(i,j) abs(m(i,j)+m(i+1,j)) / (abs(m(i,j)) + abs(m(i+1,j)));
c(2,2)
I have a range of data represented in the vector C
and have the data classes represented by the vector R
C = [1.71974522292994
1.91974522292994
2.03821656050955
2.13375796178344
2.16560509554140
2.22929936305733
2.35668789808917
2.38853503184713
2.54777070063694
2.61146496815287
2.70700636942675
2.73885350318471
2.83439490445860
2.96178343949045
3.02547770700637
3.31210191082803]
R = [1.71974522292994
2.03821104580359
2.35667686867724
2.67514269155088
2.99360851442453
3.31207433729818
3.63054016017183]
I need to do a histogram and a curve to overlap Standard Normal
z = histc(C,R); bar(R,z);
but the vector z that represents the frequency is not correct.
z = [2 4 4 4 1 1]'
on excell is so, and represents well the histogram
z = [1 1 4 4 4 1 1]'
you could suggest a solution using these two vectors?
Tnks
That's because Matlab's definition of histc:
n(k) counts the value x(i) if edges(k) <= x(i) < edges(k+1)
whereas Excel probably uses the more standard
edges(k) < x(i) <= edges(k+1)
So essentially you need to move the equal sign from below to above. You can get that either
By the trick of changing signs to both vectors and flipping the second (to keep it sorted):
>> z = histc(-C,-R(end:-1:1))
z =
1
1
4
4
4
1
1
Using the very powerful bsxfun function to directly compute the histogram with the equal sign above:
z = diff(sum(bsxfun(#le, C(:), [-inf R(:).'])));
I found it helpful to use this
z = histc(-C,-R);
the loop that I used for inside includes matrices with vectors of different length. I then filled the matrix with NaN
C and classe = [30x14] created by vectors with different lengths + NaN
[nr,nc] = size(C);
Freq = NaN*ones(nr,nc);
R = NaN*ones(nr,nc);
CC = NaN*ones(nr,nc);
I do not find a way in order to create the correct number of figures with the correct subplot.
In each figure there must be 4 subplot.
for k = 1:4
for j= 1 : nc;
R = classe(:,j);
CC = C(:,j);
FF = Freq(:,j);
R = R(~isnan(B)); % toglie i valori NaN
CC = CC(~isnan(CC));
R = sort(R,'descend');
CC = sort(C,'descend');
the line of Plot
FF = histc(-C,-R); % Calculate the J-th absolute frequencies
figure(k); <===?????
subplot(2,2,k) <=== ????????
bar(B,FF);
reassemble the matrices
if length(B)<nr
R(length(R)+1:nr)=NaN; % riempie la parte di colonna vuota
if length(CC)< nr;
C(length(CC)+1:nr)=NaN;
if length(FF)< nr;
FF(length(FF)+1:nr)=NaN;
end
end
end
classe(:,j)=R(:); % matrice classe
C(:,j)=CC(:) % matrice Elementi;
Freq(:,j)=FF(:); %Matrice Frequenze Assolute
end
the next steps concerns the in plotting of 3 figures. Each figure contains 4 subplot
I would also be able to superimpose on each histogram a standard normal curve ...
a tip?
tnks
I have a vector of length 3. i want to represent it as a matrix of dimension 4*2. ie) if the length of vector is n then matrix should be of dimension (n+1)*2. The matrix should have elements arranged as follows:
Vector= [2 3 4]
Matrix = [0 2;2 3;3 4;4 0]
You can solve your problem easily with simple operations:
vector = [2 3 4];
matrix = [0 vector; vector 0]';
' is used to transpose the matrix.
Additionally there are two useful functions in Matlab to manipulate matrices and vectors:
reshape()
repmat()
The command reshape from Matlab is the basis of my answer to your question:
B = reshape(A,m,n) returns the m-by-n matrix B whose elements are taken column-wise from A. An error results if A does not have m*n elements (from the official Matlab help).
You basically add zeros at the beginning and at the end and then have every number in the vector occur twice (if you "unfold"/reshape the matrix). So lets construct the desired matrix by reversing this description:
%set input vector
v = [2 3 4];
%"double" the numbers, v_ is my temporary storage variable
v_ = [v; v];
%align all numbers along one dimension
v_ = reshape(v_, 2*length(v), 1)
%add zeros at beginning and end
v_ = [0 v_ 0];
%procude final matrix
m = reshape(v_, length(v)+1, 2);
in short
%set input vector
v = [2 3 4];
%"double" the numbers, v_ is my temporary storage variable
%all values are aligned as row vector
%zeros are added at beginning and end
v_ = [0, v, v, 0];
%produce final matrix
m = reshape(v_, length(v)+1, 2);
I haven't checked it, since I don't have a Matlab at hand right now, but you should get the idea.
Edit
The answer by 13aumi manages this task even without the reshape command. However, you need to pay close attention to the shape of v (row- vs- column-vector).