I have a cell M with n number of cells, with each cell containing several unique numbers, something like this:
{[15 16 21 26 28 145],[2 5 8 9 15],[20 24 27],[10 11 15 8 6 258 74 1],...}
Some of these values appear in more than 1 cell. I would like to calculate the fraction of overlapping values across these cells. For instance, with the 4 cells above, I have 19 unique numbers, and 2 of them belong to more than 1 cell: 15 and 8. Thus, the fraction of overlapping cell is 2/19 = .105. Note that the number of cells in M can vary and thus the number of unique numbers in M also vary as well. Does anyone have any suggestion on how to do this efficiently? I've tried horzcat to concatenate the cells within M then used unique but didn't quite get what I want.
Using the output of the hist() function is useful here.
M = {[15 16 21 26 28 145],[2 5 8 9 15],[20 24 27],[10 11 15 8 6 258 74 1]};
% Bring data to one matrix.
M2 = cell2mat(M);
% Build a histogram from the data with a bin on each unique element. The
% first output of hist is the number of elements in each bin.
a = hist(M2,unique(M2));
% Calculate the overlap by dividing the number of elements that occur more
% than once by the total number of elements.
overlap = sum(a>1)/numel(a);
Related
I have data of integers in x = 500 X 612 matrix. I need a new variable xx in a 500 X 612 matrix but I need to apply cumsum along each row (500) across 12 column steps and applying cumsum like this 51 times --> 500 X (12 X 51) matrix. Then I need a for loop to produce 51 plots of the 500 rows and 12 columns of the cumsum time series. thank you!
I will rephrase what the question is asking to benefit those who are reading.
The OP wishes to segment a matrix into chunks by splitting up the matrix into a bunch of columns. A cumsum is applied to each row individually for each column and are then concatenated together to build a final matrix. As such, given this source matrix:
x =
1 2 3 4 5 6 7 8 9 10 11 12
13 14 15 16 17 18 19 20 21 22 23 24
Supposing that we wish to split up the matrix by columns 3, 6 and 9 and 12, we will have four chunks to work with. We do a cumsum on each of these blocks individually and piece the final result together. So the result would like the following:
xx =
1 3 6 4 9 15 7 15 24 10 21 33
13 27 42 16 33 51 19 39 60 22 45 69
First, you need to determine how many columns you want to break up the matrix into. In your case, we wish to segment the matrix into 4 chunks: Columns 1 - 3, columns 4 - 6, columns 7 - 9, and columns 10 - 12. As such, I'm going to reshape this matrix so that each column is an individual row from a chunk in this matrix. We then apply cumsum over this reshaped matrix and we then reshape it back to what you had originally.
Therefore, do this:
num_chunks = 4; %// Columns 3, 6, 9, 12
divide_point = size(x,2) / num_chunks; %// Determine how many elements are in a row for a cumsum
x_reshape = reshape(x.', divide_point, []); %// Get reshaped matrix
xy = cumsum(x_reshape); %// cumsum over all columns individually
xx = reshape(xy, size(x,2), size(x,1)).'; %// Reconstruct matrix
In the third line of code, x_reshape = reshape(x.', divide_point, []); may seem a bit daunting, but it's actually not that bad. I had to transpose the matrix first because you want to take each row of a chunk and place them into individual columns so we can perform a cumsum on each column. When you reshape something in MATLAB, it collects values column-wise and reshapes the input into an output of a specified size. Therefore, to collect the rows, we need to collect row-wise and so we must transpose this matrix. Next, divide_point tells you how many elements we have for a single row in one chunk. As such, we want to construct a matrix that is of size divide_point x N where divide_point tells you how many elements we have in a row of a chunk and N is the total number of rows over all chunks. Because I don't want to calculate how many there are (am rather lazy actually....), the [] syntax is to automatically infer this number so that we can get a reshaped matrix that respects the total number of elements in the original input. We then perform cumsum on each of these columns, and then we need to reshape this back into the original shape of the input. With this, we use reshape again on the cumsum result, but in order to get it back into the row-order that you want, we have to determine the transpose as reshape takes values in column-major order, then re-transpose that result.
We get:
xx =
1 3 6 4 9 15 7 15 24 10 21 33
13 27 42 16 33 51 19 39 60 22 45 69
In general, the total number of elements to sum over for a row needs to be evenly divisible by the total number of columns that your matrix contains. For example, given the above, if you were to try to segment this matrix into 5 chunks, you would certainly get an error as the number of rows to cumsum over is not symmetric.
As another example, let's say we wanted to break up the matrix into 6 chunks. Therefore, by setting num_chunks = 6, we get:
xx =
1 3 3 7 5 11 7 15 9 19 11 23
13 27 15 31 17 35 19 39 21 43 23 47
You can see that cumsum restarts at every second column, as we desired 6 chunks and to get 6 chunks with a matrix of 12 columns, a chunk is created at every second column.
I have a sparse matrix:
A=
(14,13) 0.5286
(15,14) 0.6781
(16,15) 0.5683
(17,16) 1.2773
(18,17) 1.0502
(19,18) 0.4966
(21,19) 0.9951
(21,20) 0.4522
(22,21) 0.8507
(23,22) 1.0727
(24,23) 0.8288
(25,24) 0.5811
(26,25) 0.8235
(28,26) 1.5128
(30,28) 0.7966
(30,29) 0.6363
(31,29) 0.8254
(32,31) 0.8573
(33,32) 1.0753
that is result of a minimum spanning tree. now I want to extract 13,14,15,...26,28,29,...33.
as seen 27 is not between numbers. so pred give: 13 14 15 16 17 18 19 21 22 23 24 25 26 28 29 30 31 32 that 20 and 33 is not.
how can I extract total of numbers that say in top?
[ii jj] = find(A);
answer = unique([ii(:); jj(:)]);
should do it.
Note that the find command with two outputs gives you the row and column index of all nonzero elements. Since you have a minimum spanning tree, each number you care about needs to occur at least once in the row or column (for example your matrix never has the number 29 in the first index, but it occurs in the second).
The unique function makes sure that each number that occurs is only represented once.
I have a matrix and i want to consider it has 4 sub matrices which are placed together. How can I find the middle element of each sub matrix when they are together?
consider the matrix below. It is built by 4 sub matrices.
1 2 3 4 5 6
7 8 9 10 11 12
13 14 15 16 17 18
19 20 21 22 23 24
25 26 27 28 29 30
31 32 33 34 35 36
I want to get their middle elements so i could have:
8, 11, 26, 29
From what I have understood this might work for you and this is a demo, so use your own parameters -
Code
%%// Input matrix
A = rand(44,44);
%%/ Number of submatrices needed
num_submat = 16;%%// 4 for your example case
%%/ Number of submatrices along row and column
num_submat1= sqrt(num_submat);
%%// Middle element indices along each direction
v1 = floor(size(A,2)/(2*num_submat1))+1:size(A,2)/(num_submat1):size(A,2);
%%// Middle elements
middle_ele = A(v1,v1)
It is always helpful to know that matrix indexing in matlab goes columnwise eg,
indOrd = [1,4,7;2,5,8;3,6,9]
where the number is the index order and not related to your example. indOrd(4) would return 4. Try to use this to find the index locations.
Assuming that each submatrix has odd size 2n+1, the coordinates of the center of one submatrix alone are [n+1, n+1]. If your have a square with M*M submatrices (M=2 in your case), the coordinates are [n+1+i*(2*n+1), n+1+j*(2*n+1)], i and j taken independently in the range 0:M-1.
Turning back to Matlab, it is now quite easy to generate all indices of the centers of the submatrices grouped in the matrix A:
n = floor(size(A,1)/(2*M));
xc = n+1+reshape(repmat(0:M-1,M,1),[],1);
yc = n+1+reshape(repmat((0:M-1)',1,M),[],1);
centers = A(yc, xc);
For even-sized submatrix, you have to choose which element is the center, the modification is quite easy to do then.
I have a vector of values which represent an index of a row to be removed in some matrix M (an image). There's only one row value per column in this vector (i.e. if the image is 128 x 500, my vector contains 500 values).
I'm pretty new to MATLAB so I'm unsure if there's a more efficient way of removing a single pixel (row,col value) from a matrix so I've come here to ask that.
I was thinking of making a new matrix with one less row, looping through each column up until I find the row whose value I wish to remove, and "shift" the column up by one and then move onto the next column to do the same.
Is there a better way?
Thanks
Yes, there is a solution which avoids loops and is thus faster to write and to execute. It makes use of linear indexing, and exploits the fact that you can remove a matrix entry by assigning it an empty value ([]):
% Example data matrix:
M = [1 5 9 13 17
2 6 10 14 18
3 7 11 15 19
4 8 12 16 20];
% Example vector of rows to be removed for each column:
vector = [2 3 4 1 3];
[r c] = size(M);
ind = sub2ind([r c],vector,1:c);
M(ind) = [];
M = reshape(M,r-1,c);
This gives the result:
>> M =
1 5 9 14 17
3 6 10 15 18
4 8 11 16 20
I want to calculate the sum of the elements surrounding a given element in a matrix. So far, I have written these lines of code:
for i=1:m,
rij(1:n)=0
for j=1:n,
alive = tijdelijk(i-1,j)+tijdelijk(i+1,j)+tijdelijk(i-1,j-1)+tijdelijk(i+1,j-1)+tijdelijk(i,j+1)+tijdelijk(i,j-1)+tijdelijk(i-1,j+1)+tijdelijk(i+1,j+1)
This results in an error because, for example, i-1 becomes zero for i=1. Anyone got an idea how to do this without getting this error?
You can sum the elements via filtering. conv2 can be used for this manner.
Let me give an example. I create a sample matrix
>> A = reshape(1:20, 4, 5)
A =
1 5 9 13 17
2 6 10 14 18
3 7 11 15 19
4 8 12 16 20
Then, I create a filter. The filter is like a mask where you put the center on the current cell and the locations corresponding to the 1's on the filter are summed. For eight-connected neighbor case, the filter should be as follows:
>> B = [1 1 1; 1 0 1; 1 1 1]
B =
1 1 1
1 0 1
1 1 1
Then, you simply convolve the matrix with this small matrix.
>> conv2(A, B, 'same')
ans =
13 28 48 68 45
22 48 80 112 78
27 56 88 120 83
18 37 57 77 50
If you want four-connected neighbors, you can make the corners of your filter 0. Similarly, you can design any filter for your purpose, such as for averaging all neighbors instead of summing them.
For details, please see the convolution article in Wikipedia.
Two possibilities : change the limits of the loops to i=k:(m-k) and j=k:(n-k) or use blkproc
ex :
compute the 2-D DCT of each 8-by-8 block
I = imread('cameraman.tif');
fun = #dct2;
J = blkproc(I,[8 8],fun);
imagesc(J), colormap(hot)
There are lots of things you can do at the edges. Which you do depends very specifically on your problem and is different from usage case to usage case. Typical things to do:
If (i-1) or (i+1) is out of range, then just ignore that element. This is equivalent to zero padding the matrix with zeros around the outside and adjusting the loop limits accordingly
Wrap around the edges. In other words, for an MxN matrix, if (i-1) takes you to 0 then instead of taking element (i-1, j) = (0, j) you take element (M, j).
Since your code mentions "your teacher" I'd guess that you can ask what should happen at the edges (or working it out in a sensible manner may well be part of the task!!).