My program creates a matrix whose cell values in multiple rows are the same in the corresponding column. I want to delete some rows that have 0 more than one. To clarify, my matrix has the following form,
A=[ 1 1 1 0 0 1 1 1; 1 0 0 1 1 1 1 1 1; 1 1 1 1 1 1 1 0; 1 1 1 1 0 1 1 1
1 1 0 1 0 0 1 1 ]
and I want to delete all the columns that are in the first, second and fifth rows because the number 0 is 2
or more left in the matrix of rows that are in the third and fourth rows because they have 0 one in each row.
The result should be the following matrix:
A=[ 1 1 1 1 1 1 1 1 0; 1 1 1 1 0 1 1 1 ]
i write this code for your algorithm than work correctly:
% Input Matrix
A = [1 1 1 0 0 1 1 1;1 0 0 1 1 1 1 1; 1 1 1 1 1 1 1 0;1 1 1 1 0 1 1 1;1 1 0 1 0 0 1 1 ];
% find number of rows and cols
[num_rows, num_cols] = size(A);
% Itrate on each row and find rows that have less than 2 zeros
selected_rows = [];
idx = 1;
for i=1:num_rows
num_zero = sum(A(i, 1:end) == 0);
if num_zero < 2
selected_rows(idx) = i;
idx = idx+1;
end
end
% return result matrix
result = [];
for i=1:length(selected_rows)
result = [result; A(selected_rows(i), 1:end)];
end
disp(result)
Related
I've been struggling with this for a bit now. I have a small matrix s for example and a bigger matrix B as shown below.
B =
0 0 0 0 0 0 1 1
1 1 0 0 1 0 1 1
1 1 0 1 0 0 1 1
1 1 1 0 0 0 1 0
0 0 1 1 1 0 0 1
0 0 0 1 1 1 1 1
1 1 1 0 0 0 1 0
0 1 1 0 1 1 0 0
s =
1 1
1 1
What I want to do is iterate through B with s and compare the values. If all the values in s equal the values in B (the small section of B), then the answer is 1, if not then 0.
The 1's and 0's would be placed in a matrix as well.
This is what I've done so far but unfortunately, it doesn't iterate step by step and doesn't create a matrix either.
s = ones(2,2)
B = randi([0 1],8,8)
f = zeros(size(B))
[M,N]=size(B); % the larger array
[m,n]=size(s); % and the smaller...
for i=1:M/m-(m-1)
for j=1:N/n-(n-1)
if all(s==B(i:i+m-1,j:j+n-1))
disp("1")
else
disp("0")
end
end
end
Any help would be appreciated!
The following code works on the examples you supplied, I haven't tested it on anything else, and it will not work if the dimensions of the smaller matrix are not factors of the dimensions of the larger matrix, but you didn't indicate that it needed to do that in your description.
B =[0 0 0 0 0 0 1 1
1 1 0 0 1 0 1 1
1 1 0 1 0 0 1 1
1 1 1 0 0 0 1 0
0 0 1 1 1 0 0 1
0 0 0 1 1 1 1 1
1 1 1 0 0 0 1 0
0 1 1 0 1 1 0 0];
S =[1 1
1 1];
%check if array meets size requirements
numRowB = size(B,1);
numRowS = size(S,1);
numColB = size(B,2);
numColS = size(S,2);
%get loop multiples
incRows = numRowB/numRowS;
incCols = numColB/numColS;
%create output array
result = zeros(incRows, incCols);
%create rows and colums indices
rowsPull = 1:numRowS:numRowB;
colsPull = 1:numColS:numColB;
%iterate
for i= 1:incRows
for j= 1:incCols
result(i,j) = isequal(B(rowsPull(i):rowsPull(i)+numRowS-1, colsPull(j):colsPull(j)+numColS-1),S);
end
end
%print the resulting array
disp(result)
I have 2 matrices
Matrix A = [7 3 5 2 8 4 1 6 9;
5 2 6 1 4 3 9 7 8;
9 1 4 5 2 6 3 6 7;
4 8 1 6 3 7 2 9 5;
6 1 7 2 8 4 5 9 3]
Matrix B = [1 0 0 0 0 0 0 0 0;
0 1 1 0 0 0 0 0 0;
0 0 0 0 0 0 0 1 0;
0 0 0 1 0 1 0 0 0;
0 0 0 0 0 0 0 0 1]
Matrix A and B are already defined.
Here each column can't have more than 1 what i want to do is that if when i do sum for Matrix B if i found 0 in it i have to add 1's in the places of the zero's but in certain places. In each row the 1's have to be placed in certain groups. For example if a 1 is placed in column 1, then it can be placed as well in column 2 or 3 only. It can't be placed anywhere else. If in another row it is placed in column 5, then it can be placed in column 4 or 6 only and so on. It's like group of 3. Each 3 columns are together.
To be more clear:
Here the sum of matrix B is [1 1 1 1 0 1 0 1 1]. The zeros here are placed in column 5 and 7 and i want to add 1 putting in mind where the 1 is going to be placed in the matrix. So in this example the 1 of column 5 can only be placed in row 4 as the 1's in this row are placed in column 4 and 6. The 1 of column 7 can be placed in row 5 or row 3. If we have choice between 2 rows then the 1 will be placed in the placed of the higher number of Matrix A.
The 1's have to be placed in groups; columns 1, 2 and 3 are together, columns 4,5 and 6 are together and columns 7, 8 and 9 are together. so if the 1 is placed in 1 column of the group then it can't be placed in any other place.
Let me simplify it if we have an array like this [0 0 0 0 0 0 0 1 1] This array has 3 categories, columns 1,2 and 3 are 1st category, columns 4,5 and 6 are 2nd category and so on. here i want to place a 1 so that the 3rd category won't have a zero element. This is what i want to do briefly but with a whole matrix with all the categories.
so here the output will be =
[1 0 0 0 0 0 0 0 0;
0 1 1 0 0 0 0 0 0;
0 0 0 0 0 0 0 1 0;
0 0 0 1 1 1 0 0 0;
0 0 0 0 0 0 1 0 1]
This code was tried but it doesn't give the required output as the 1 was placed in the 1st row not in the place where it has to be (the category that it should be in).
sum_cols_B = sum(B); % Sum of Matrix B (dim 1)
[~, idx] = find(sum_cols_B == 0); % Get indices where sum == 0
% Using loop to go through the indices (where sum = 0)
for ii = idx
B(1,ii) = 1; % Insert 1 in the first position of that
end % column in Matrix B
Ask me if the question is still not clear.!
Here's an updated loop that will add the missing 1's:
sum_cols_B = sum(B);
[~, idx] = find(sum_cols_B == 0);
group_size = 3;
for ii = idx
% Calculate the starting column of the group for column ii
% There are (ii-1)/group_size groups
% Add 1 for 1-based indexing
group_start = floor((ii-1)/group_size)*group_size + 1;
% Determine which rows in the current group have nonzero values
group_mask = sum(B(:,group_start:group_start+group_size-1), 2) > 0;
% Find the row number of the max in A column ii corresponding to mask
[~,rownum] = max(A(:,ii).*group_mask);
% The value in column ii of B should have a 1 inserted
% at the row containing the max in A
B(rownum,ii) = 1;
end
Results for B above are:
B =
1 0 0 0 0 0 0 0 0
0 1 1 0 0 0 0 0 0
0 0 0 0 0 0 0 1 0
0 0 0 1 1 1 0 0 0
0 0 0 0 0 0 1 0 1
B = [1 0 0 0 0 0 0 0 0;...
0 1 1 0 0 0 0 0 0;...
0 0 0 0 0 0 0 1 0;...
0 0 0 1 0 1 0 0 0;...
0 0 0 0 0 0 0 0 1]; % Matrix - using this as an example
sum_cols_B = sum(B); % Sum of Matrix B (dim 1)
[~, idx] = find(sum_cols_B == 0); % Get indices where sum == 0
% Using loop to go through the indices (where sum = 0)
for ii = idx
B(1,ii) = 1; % Insert 1 in the first position of that
end % column in Matrix B
I need to replace the repeated elements in column of a matrix as 0's and delete the rows which has all 0's. If my matrix is like this means.
Input =
1 0 0 1
0 1 0 1
0 0 1 1
1 1 1 1
My expected output should be like this
Output =
1 0 0 1
0 1 0 0
0 0 1 0
0 0 0 0 ---> this row should be get deleted in this case
This doesn't work for my problem
c = [ 1 1 0 1 0 1 1 1 0 1 1 0];
[c, ic] = unique(a, 'first');
c(~ismember(1:length(a),ic)) = 0;
You can use logical indexing and cumsum:
A = [1 0 0 1;
0 1 0 1;
0 0 1 1;
1 1 1 1];
ind = cumsum(A); %cumulative sum (by column)
A(ind>1) = 0;
A(sum(A')==0,:)=[]
I'd like to create a Matrix in MATLAB where:
The First row consists of a random arrangement of 0s and 1s, split evenly (i.e. 50-50).
The Second row randomly assigns zeros to 50% of the 0s and 1s in the first row, and ones to the remaining 50%.
The Third row randomly assigns zeros to 50% of the 0s and 1s in the second row, and ones to the remaining 50%.
Non-randomized Example:
0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1
0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1
0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1
Any suggestions?
A solution based on checking whether numbers are bigger or smaller than the median. As long as the number of columns tested is even, exactly half of a set of random doubles will be bigger than the median, and half will be smaller. This guarantees that there's exactly 50% of bits get flipped.
nRows = 3;
nCols = 16; %# divisible by 4
%# seed the array
%# assume that the numbers in each row are unique (very, very likely)
array = rand(nRows,nCols);
out = false(nRows,nCols);
%# first row is special
out(1,:) = array(1,:) > median(array(1,:));
%# for the rest of the row, check median for the zeros/ones in the previous row
for iRow = 2:nRows
zeroIdx = out(iRow-1,:) == 0;
%# > or < do not matter, both will replace zeros/ones
%# and replace with exactly half zeros and half ones
out(iRow,zeroIdx) = array(iRow,zeroIdx) > median(array(iRow,zeroIdx));
out(iRow,~zeroIdx) = array(iRow,~zeroIdx) > median(array(iRow,~zeroIdx));
end
I'd offer a short bsxfun solution:
%// number of divisions
n = 4;
%// unshuffled matrix like in your example
unshuffled = bsxfun(#(a,b) mod(a,2*b) > b-1, meshgrid(1:n^2,1:n) - 1, (2.^((n-1):-1:0)).') %'
%// shuffle columns
shuffled = unshuffled(:,randperm(n^2))
unshuffled =
0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1
0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1
0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1
0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
shuffled =
1 0 1 1 0 1 0 1 1 1 1 0 0 0 0 0
1 1 1 0 0 1 1 0 1 0 0 0 1 0 1 0
1 0 0 1 0 0 0 0 1 1 0 1 1 0 1 1
1 1 1 1 0 0 0 0 0 0 1 0 0 1 1 1
First you need to create the unshuffled matrix, which can be done by comparing the matrix generated by meshgrid(1:n^2,1:n) with a row dependent modulus. Finally you just need to shuffle the columns.
If you have the Statistics Toolbox, you can do it very easily with randsample:
M = 3; %// number of rows
N = 16; %// number of columns. Should be multiple of 4, according to problem definition
result = zeros(M,N); %// preallocate and initiallize to zeros
result(1, randsample(1:N,N/2)) = 1; %// first row: half values set to one, half to zero
for m = 2:M
result(m, :) = result(m-1, :); %// initiallize row m equal to row m-1
result(m, randsample(find(result(m-1,:)), N/4)) = 0; %// change half of ones
result(m, randsample(find(~result(m-1,:)), N/4)) = 1; %// change half of zeros
end
Example result:
result =
0 1 0 1 1 0 0 0 0 1 0 1 1 0 1 1
1 1 0 0 0 1 1 1 0 1 0 1 0 0 0 1
1 0 0 0 1 0 0 1 0 1 1 0 1 1 0 1
A solution using randperm:
nrows = 3;
ncols = 16;
M = zeros(nrows,ncols);
%// seed the first row
M(1,1:ncols/2) = 1;
M(1,:) = M(1,randperm(ncols));
for r = 2:nrows
%// Find ncols/4 random between 1 and ncols/2. These will be used to index half of the previous rows 1 elements and set them to one
idx = randperm(ncols/2);
idx1 = idx(1:ncols/4);
%// Do the same thing again, but this time it will be used for the 0 elements of the previous row
idx = randperm(ncols/2);
idx0 = idx(1:ncols/4);
idx_prev1 = find(M(r-1,:)); %// Find where the 1 elements were in the last row
idx_prev0 = find(~M(r-1,:)); %// Find where the 0 elements were in the last row
M(r,idx_prev1(idx1))=1; %// Set half of the previous rows 1 elements in this row to 1
M(r,idx_prev0(idx0))=1; %// Set half of the previous rows 0 elements in this row to 1
end
Let's have a M = [10 x 4 x 12] matrix. As example I take the M(:,:,4):
val(:,:,4) =
0 0 1 0
0 1 1 1
0 0 0 1
1 1 1 1
1 1 0 1
0 1 1 1
1 1 1 1
1 1 1 1
0 0 1 1
0 0 1 1
How can I obtain this:
val(:,:,4) =
0 0 3 0
0 2 2 2
0 0 0 4
1 1 1 1
1 1 0 1
0 2 2 2
1 1 1 1
1 1 1 1
0 0 3 3
0 0 3 3
If I have 1 in the first column then all the subsequent 1's should be 1.
If I have 0 in the first column but 1 in the second, all the subsequent 1's should be 2.
If I have 0 in the first and second column but 1 in the third then all the subsequent 1's should be 3.
If I have 0 in the first 3 columns but 1 in the forth then this one should be four.
Note: The logical matrix M is constructed:
Tab = [reshape(Avg_1step.',10,1,[]) reshape(Avg_2step.',10,1,[]) ...
reshape(Avg_4step.',10,1,[]) reshape(Avg_6step.',10,1,[])];
M = Tab>=repmat([20 40 60 80],10,1,size(Tab,3));
This is a very simple approach that works for both 2D and 3D matrices.
%// Find the column index of the first element in each "slice".
[~, idx] = max(val,[],2);
%// Multiply the column index with each row of the initial matrix
bsxfun(#times, val, idx);
This could be one approach -
%// Concatenate input array along dim3 to create a 2D array for easy work ahead
M2d = reshape(permute(M,[1 3 2]),size(M,1)*size(M,3),[]);
%// Find matches for each case, index into each matching row and
%// elementwise multiply all elements with the corresponding multiplying
%// factor of 2 or 3 or 4 and thus obtain the desired output but as 2D array
%// NOTE: Case 1 would not change any value, so it was skipped.
case2m = all(bsxfun(#eq,M2d(:,1:2),[0 1]),2);
M2d(case2m,:) = bsxfun(#times,M2d(case2m,:),2);
case3m = all(bsxfun(#eq,M2d(:,1:3),[0 0 1]),2);
M2d(case3m,:) = bsxfun(#times,M2d(case3m,:),3);
case4m = all(bsxfun(#eq,M2d(:,1:4),[0 0 0 1]),2);
M2d(case4m,:) = bsxfun(#times,M2d(case4m,:),4);
%// Cut the 2D array thus obtained at every size(a,1) to give us back a 3D
%// array version of the expected values
Mout = permute(reshape(M2d,size(M,1),size(M,3),[]),[1 3 2])
Code run with a random 6 x 4 x 2 sized input array -
M(:,:,1) =
1 1 0 1
1 0 1 1
1 0 0 1
0 0 1 1
1 0 0 0
1 0 1 1
M(:,:,2) =
0 1 0 1
1 1 0 0
1 1 0 0
0 0 1 1
0 0 0 1
0 0 1 0
Mout(:,:,1) =
1 1 0 1
1 0 1 1
1 0 0 1
0 0 3 3
1 0 0 0
1 0 1 1
Mout(:,:,2) =
0 2 0 2
1 1 0 0
1 1 0 0
0 0 3 3
0 0 0 4
0 0 3 0