I am making a 3 dimensional array, using matlab, that progresses according to a pattern. Athough I could write out the Array manually I am sure there is a quicker way to do it.
multiArray = cat(3,...
[1+randn(4,3); 1*randn(4,3)],...
[2+randn(4,3); 2*randn(4,3)],...
[3+randn(4,3); 3*randn(4,3)]);
If I want to make the above array to be 8x3x25 then the last line would be
[25+randn(4,3); 25*randn(4,3)]
But how can I make such an array without going through all the tedious intervening steps?
While mikkola basically got the solution, there is no need to shift dimensions at the end.
s=[4,3,25];
it=reshape(1:s(3),1,1,[]);
out = [bsxfun(#plus , it, randn(s));...
bsxfun(#times, it, randn(s))];
Here's a possible way using bsxfun.
%// 25 x 4 x 3 with elements for i + randn(4,3)
P = bsxfun(#plus, (1:25)', randn(25,4,3));
%// 25 x 4 x 3 with elements for i * randn(4,3)
T = bsxfun(#times, (1:25)', randn(25,4,3));
%// Concatenate and shift dimensions to get desired size output
multiArray = shiftdim([P T], 1);
If you don't mind taking things to 4D for efficiency purposes -
N = 25; %// Number of 3D slices
out = randn(4,2,3,N);
out(:,1,:,:) = bsxfun(#plus,permute(1:N,[1 4 3 2]),out(:,1,:,:));
out(:,2,:,:) = bsxfun(#times,permute(1:N,[1 4 3 2]),out(:,2,:,:));
out = reshape(out,8,3,N);
To legitimize the solution, let's start off with an input of A = randn(8,3,N) and initialize the output out with it. Also, let's take number of 3D slices as a small number, so say N = 3.
Thus,
>> N = 3;
A = randn(8,3,N);
out = reshape(A,[4 2 3 N]); %// This replaces "out = randn(4,2,3,N)"
Next up, we run the code that will change out -
>> out(:,1,:,:) = bsxfun(#plus,permute(1:N,[1 4 3 2]),out(:,1,:,:));
out(:,2,:,:) = bsxfun(#times,permute(1:N,[1 4 3 2]),out(:,2,:,:));
out = reshape(out,8,3,N);
Now, start validating per 3D slice -
>> out(1:4,:,1) - A(1:4,:,1)
ans =
1 1 1
1 1 1
1 1 1
1 1 1
>> out(1:4,:,2) - A(1:4,:,2)
ans =
2 2 2
2 2 2
2 2 2
2 2 2
>> out(1:4,:,3) - A(1:4,:,3)
ans =
3 3 3
3 3 3
3 3 3
3 3 3
>> out(5:end,:,1)./A(5:end,:,1)
ans =
1 1 1
1 1 1
1 1 1
1 1 1
>> out(5:end,:,2)./A(5:end,:,2)
ans =
2 2 2
2 2 2
2 2 2
2 2 2
>> out(5:end,:,3)./A(5:end,:,3)
ans =
3 3 3
3 3 3
3 3 3
3 3 3
Related
This question already has answers here:
Generate a matrix containing all combinations of elements taken from n vectors
(4 answers)
Closed 6 years ago.
I'm trying to do the following in a general way:
x = {0:1, 2:3, 4:6};
[a,b,c] = ndgrid(x{:});
Res = [a(:), b(:), c(:)]
Res =
0 2 4
1 2 4
0 3 4
1 3 4
0 2 5
1 2 5
0 3 5
1 3 5
0 2 6
1 2 6
0 3 6
1 3 6
I believe I have to start the following way, but I can't figure out how to continue:
cell_grid = cell(1,numel(x));
[cell_grid{:}] = ndgrid(x{:});
[cell_grid{:}]
ans =
ans(:,:,1) =
0 0 2 3 4 4
1 1 2 3 4 4
ans(:,:,2) =
0 0 2 3 5 5
1 1 2 3 5 5
ans(:,:,3) =
0 0 2 3 6 6
1 1 2 3 6 6
I can solve this in many ways for the case with three variables [a, b, c], both with and without loops, but I start to struggle when I get more vectors. Reshaping it directly will not give the correct result, and mixing reshape with permute becomes really hard when I have arbitrary number of dimensions.
Can you think of a clever way to do this that scales to 3-30 vectors in x?
You can use cellfun to flatten each of the cell array elements and then concatenate them along the second dimension.
tmp = cellfun(#(x)x(:), cell_grid, 'uniformoutput', false);
out = cat(2, tmp{:})
Alternately, you could avoid cellfun and concatenate them along the dimension that is one higher than your dimension of each cell_grid member (i.e. numel(x) + 1). Then reshape to flatten all dimensions but the last one you just concatenated along.
out = reshape(cat(numel(x) + 1, cell_grid{:}), [], numel(x));
I have been trying to write a code where the users enters two numbers in order to get 2 columns. It is very hard to explain by words what I am trying to achieve so here is an example:
If the user inputs a = 1 and b = 1, the following table should be created:
ans =
1 1
If the user inputs a = 2 and b = 2:
ans =
1 1
1 2
2 1
2 2
If the user inputs a = 2 and b = 5:
ans =
1 1
1 2
1 3
1 4
1 5
2 1
2 2
2 3
2 4
2 5
For other values of a and b, the matrix should be constructed according to the above shown sequence.
This can be achieved straight-forward by the use of repelem and repmat:
[repelem((1:a).',b),repmat((1:b).',a,1)]
A more elegant way is using meshgrid and reshape it after:
[A,B] = meshgrid(1:a,1:b);
[A(:),B(:)]
Let's create an anonymous function and test the first approach:
>> fun = #(a,b) [repelem((1:a).',b),repmat((1:b).',a,1)];
>> fun(1,1)
ans =
1 1
>> fun(2,2)
ans =
1 1
1 2
2 1
2 2
>> fun(2,5)
ans =
1 1
1 2
1 3
1 4
1 5
2 1
2 2
2 3
2 4
2 5
Here is an alternative way
for example a = 2 and b = 5
A(1:b*a,1) = reshape(mtimes((1:a).',ones(1,b)).',1,b*a)
A(1:b*a,2) = reshape(mtimes((1:b).',ones(1,a)),1,b*a)
A =
1 1
1 2
1 3
1 4
1 5
2 1
2 2
2 3
2 4
2 5
There is just one logic, in the code below you define a matrix of row size and a and column size b
>> mtimes((1:a).',ones(1,b))
ans =
1 1 1 1 1
2 2 2 2 2
and the next step simply reshapes the matrix column wise for a and row wise for b by taking a transpose
A(1:b*a,1) = reshape(mtimes((1:a).',ones(1,b)).',1,b*a)
A(1:b*a,2) = reshape(mtimes((1:b).',ones(1,a)),1,b*a)
Supose there is a Matrix
A =
1 3 2 4
4 2 5 8
6 1 4 9
and I have a Vector containing the "class" of each column of this matrix for example
v = [1 , 1 , 2 , 3]
How can I sum the columns of the matrix to a new matrix as column vectors each to the column of their class? In this example columns 1 and 2 of A would added to the first column of the new matrix, column 2 to the 3 to the 2nd, column 4 the the 3rd.
Like
SUM =
4 2 4
6 5 8
7 4 9
Is this possible without loops?
One of the perfect scenarios to combine the powers of accumarray and bsxfun -
%// Since we are to accumulate columns, first step would be to transpose A
At = A.' %//'
%// Create a vector of linear IDs for use with ACCUMARRAY later on
idx = bsxfun(#plus,v(:),[0:size(A,1)-1]*max(v))
%// Use ACCUMARRAY to accumulate rows from At, i.e. columns from A based on the IDs
out = reshape(accumarray(idx(:),At(:)),[],size(A,1)).'
Sample run -
A =
1 3 2 4 6 0
4 2 5 8 9 2
6 1 4 9 8 9
v =
1 1 2 3 3 2
out =
4 2 10
6 7 17
7 13 17
An alternative with accumarray in 2D. Generate a grid with the vector v and then apply accumarray:
A = A.';
v = [1 1 2 3];
[X, Y] = ndgrid(v,1:size(A,2));
Here X and Y look like this:
X =
1 1 1
1 1 1
2 2 2
3 3 3
Y =
1 2 3
1 2 3
1 2 3
1 2 3
Then apply accumarray:
B=accumarray([X(:) Y(:)],A(:)),
SUM = B.'
SUM =
4 2 4
6 5 8
7 4 9
As you see, using [X(:) Y(:)] create the following array:
ans =
1 1
1 1
2 1
3 1
1 2
1 2
2 2
3 2
1 3
1 3
2 3
3 3
in which the vector v containing the "class" is replicated 3 times since there are 3 unique classes that are to be summed up together.
EDIT:
As pointed out by knedlsepp you can get rid of the transpose to A and B like so:
[X2, Y2] = ndgrid(1:size(A,1),v);
B = accumarray([X2(:) Y2(:)],A(:))
which ends up doing the same. I find it a bit more easier to visualize with the transposes but that gives the same result.
How about a one-liner?
result = full(sparse(repmat(v,size(A,1),1), repmat((1:size(A,1)).',1,size(A,2)), A));
Don't optimize prematurely!
The for loop performs fine for your problem:
out = zeros(size(A,1), max(v));
for i = 1:numel(v)
out(:,v(i)) = out(:,v(i)) + A(:,i);
end
BTW: With fine, I mean: fast, fast, fast!
I am trying to flip certain parts of a matrix. I can explain better by example. Let's say that I have a matrix
M = [ 1 3 6;
1 2 4;
1 7 1;
2 9 0;
2 8 3;
2 4 2;
2 3 1;
3 6 5;
3 4 5;
3 1 9;
4 2 4;
4 8 6 ]
What I'd like to do here is take any rows with an even number in the first column, and flip the third column elements. The end result would look like this:
1 3 6
1 2 4
1 7 1
2 9 1 *
2 8 2 *
2 4 3 *
2 3 0 *
3 6 5
3 4 5
3 1 9
4 2 6 *
4 8 4 *
Note the rows marked with a star have had the elements of the third column flipped upside-down. The problem I'm having is going through each row like in a for-loop you cannot flip an entire set of rows.
Thanks in advance for any help.
Another time accumarray is the way to go:
A =[ 1 3 6 ;
1 2 4 ;
1 7 1 ;
2 9 0 ;
2 8 3 ;
2 4 2 ;
2 3 1 ;
3 6 5 ;
3 4 5 ;
3 1 9 ;
4 2 4 ;
4 8 6 ]
C = accumarray(A(:,1),A(:,3),[],#(x) {flipud(x)} ); %// get groups according to
%// first column and flip it
C = vertcat(C{:}); %// cell array returned,
%// transform to matrix
mask = ~mod(A(:,1),2); %// mask for even numbers
A(mask,3) = C(mask); %// replace masked values of 3rd column with flipped ones
returns:
A =
1 3 6
1 2 4
1 7 1
2 9 1
2 8 2
2 4 3
2 3 0
3 6 5
3 4 5
3 1 9
4 2 6
4 8 4
Certainly slower, but just for fun in two lines:
C = accumarray(A(:,1),A(:,3),[],#(x) {flipud(x)} );
A(~mod(A(:,1),2),3) = getfield( vertcat(C{:}), {~mod(A(:,1),2)});
%// well no, I won't explain it...
Edit: I assumed your first column just contains integers!
I would suggest you break the problem down into stages, something like so:
Identify blocks you wish to flip
Extract them
Flip them
Replace them
You can identify a set of even numbers using the unique and mod functions, then use a for loop over them and use logical indexing to pull/replace the blocks.
Here, try this
a = magic(5); % Some data in a 5x5 matrix
b = 1:numel(a); % Indices of <a>
Rearrange b however you want, then do a=a(b) to reassign a based on the reassigned indices of b. For example, the following code
disp(a(b));
would just return the elements of a in their original order. For your application this code should work:
a = <your matrix data>
b = 1:numel(a);
b = [b(1:27) fliplr(b(28:31)) b(32:34) fliplr(b(35:36))] % Change this part
a = reshape(a(b),size(a))
You should change b based on whatever you need it to do.
In Matlab, I'm trying to transform a matrix A to another matrix B such that B's columns are made up of square submatrices of A. For example, if A is:
A = [1 1 2 2
1 1 2 2
3 3 4 4
3 3 4 4];
I'd like B to be:
B = [1 2 3 4
1 2 3 4
1 2 3 4
1 2 3 4]
A could be, say 16-by-16, and constructing B from 4-by-4 squares would result in B being 4-by-64.
Is there an efficient way to do this using reshape in combination with some other commands? Or some other approach? I am currently iterating in a loop, which is very slow with a large number of large source matrices.
Assume your matrix is a bit more general, and made of 3x2 blocks:
A = [1 1 2 2
1 1 2 2
1 1 2 2
3 3 4 4
3 3 4 4
3 3 4 4
5 5 6 6
5 5 6 6
5 5 6 6];
b = [3 2];
szA = size(A);
Transpose, reshape, permute, reshape.
nb = prod(szA./b); % Number of blocks
nelb = prod(b); % Number of elements per block
out1 = reshape(permute(reshape(A',szA(2),b(1),szA(1)/b(1)),[2,1,3]),nelb,nb)
Alternatively, slower and memory intensive but more readable:
d1 = repmat(b(1),1,szA(1)/b(1));
d2 = repmat(b(2),1,szA(2)/b(2));
out = reshape(mat2cell(A,d1,d2)',1,nelb);
out = reshape([out{:}],nelb,nb)
Now, if the blocks are square, simply set b = [2,2] or b = [3,3], etc..., or simplify the general formulation removing indexing of b and prod.