I'm performing input-output calculations in Octave. I have several matrices/vectors in the formula:
F = f' * (I-A)^-1 * Y
All vectors probably contain zeroes. I would like to omit them from the calculation and just return 0 instead. Any help would be greatly appreciated!
Miranda
What do you mean when you say "omit them"?
If you want to remove zeros from a vector you can do this:
octave:1> x=[1,2,0,3,4,0,5];
octave:2> x(find(x==0))=[]
x =
1 2 3 4 5
The logic is: x==0 will test each element of x (in this case the test is if it equals zero) and will return a vector of 0's and 1's (0 if the test is false for that element and 1 otherwise)
So:
octave:1> x=[1,2,0,3,4,0,5];
octave:2> x==0
ans =
0 0 1 0 0 1 0
The find() function will return the index value of any non-zero element of it's argument, hence:
octave:3> find(x==0)
ans =
3 6
And then you are just indexing and removing when you do something like:
octave:5> x([3, 6]) = []
x =
1 2 3 4 5
But instead you do it with the output of the find() function (which is the vector [3,6] in this case)
You can do the same for matrices:
octave:7> A = [1,2,0;4,5,0]
A =
1 2 0
4 5 0
octave:8> A(find(A==0))=[]
A =
1
4
2
5
Then use the reshape() function to turn it back into a matrix.
Related
Using Julia, I'd like to determine if a row is located in a matrix and (if applicable) where in the matrix the row is located. For example, in Matlab this can done with ismember:
a = [1 2 3];
B = [3 1 2; 2 1 3; 1 2 3; 2 3 1]
B =
3 1 2
2 1 3
1 2 3
2 3 1
ismember(B, a, 'rows')
ans =
0
0
1
0
From this, we can see a is located in row 3 of B. Is there a similar function to accomplish this in Julia?
You can also make use of array broadcasting by simply testing for equality (.==) without the use of comprehensions:
all(B .== a, dims=2)
Which gives you:
4x1 BitMatrix:
0
0
1
0
You can then use findall on this array:
findall(all(B .== a, 2))
However, this gives you a vector of CartesianIndex objects:
1-element Vector{CartesianIndex{2}}:
CartesianIndex(3, 1)
So if you expect to find multiple rows with the value defined in a you can either:
simplify this Vector by taking only the row index from each CartesianIndex:
[cart_idx[1] for cart_idx in findall(all(B .== a, 2))]
or pass one dimensional BitMatrix to findall (as suggested by Shep Bryan in the comment):
findall(all(B .== a, dims=2)[:, 1])
Either way you get an integer vector of column indices:
1-element Vector{Int64}:
3
Another pattern is using array comprehension:
julia> Bool[ a == B[i,:] for i=1:size(B,1) ]
4-element Array{Bool,1}:
false
false
true
false
julia> Int[ a == B[i,:] for i=1:size(B,1) ]
4-element Array{Int64,1}:
0
0
1
0
how about:
matchrow(a,B) = findfirst(i->all(j->a[j] == B[i,j],1:size(B,2)),1:size(B,1))
returns 0 when no matching row, or first row number when there is one.
matchrow(a,B)
3
should be as "fast as possible" and pretty simple too.
Though Julia doesn't have a built-in function, its easy enough as a one-liner.
a = [1 2 3];
B = [3 1 2; 2 1 3; 1 2 3; 2 3 1]
ismember(mat, x, dims) = mapslices(elem -> elem == vec(x), mat, dims)
ismember(B, a, 2) # Returns booleans instead of ints
I'm trying to create a function that is able to go through a row vector and output the possible combinations of an n choose k without recursion.
For example: 3 choose 2 on [a,b,c] outputs [a,b; a,c; b,c]
I found this: How to loop through all the combinations of e.g. 48 choose 5 which shows how to do it for a fixed n choose k and this: https://codereview.stackexchange.com/questions/7001/generating-all-combinations-of-an-array which shows how to get all possible combinations. Using the latter code, I managed to make a very simple and inefficient function in matlab which returned the result:
function [ combi ] = NCK(x,k)
%x - row vector of inputs
%k - number of elements in the combinations
combi = [];
letLen = 2^length(x);
for i = 0:letLen-1
temp=[0];
a=1;
for j=0:length(x)-1
if (bitand(i,2^j))
temp(k) = x(j+1);
a=a+1;
end
end
if (nnz(temp) == k)
combi=[combi; derp];
end
end
combi = sortrows(combi);
end
This works well for very small vectors, but I need this to be able to work with vectors of at least 50 in length. I've found many examples of how to do this recursively, but is there an efficient way to do this without recursion and still be able to do variable sized vectors and ks?
Here's a simple function that will take a permutation of k ones and n-k zeros and return the next combination of nchoosek. It's completely independent of the values of n and k, taking the values directly from the input array.
function [nextc] = nextComb(oldc)
nextc = [];
o = find(oldc, 1); %// find the first one
z = find(~oldc(o+1:end), 1) + o; %// find the first zero *after* the first one
if length(z) > 0
nextc = oldc;
nextc(1:z-1) = 0;
nextc(z) = 1; %// make the first zero a one
nextc(1:nnz(oldc(1:z-2))) = 1; %// move previous ones to the beginning
else
nextc = zeros(size(oldc));
nextc(1:nnz(oldc)) = 1; %// start over
end
end
(Note that the else clause is only necessary if you want the combinations to wrap around from the last combination to the first.)
If you call this function with, for example:
A = [1 1 1 1 1 0 1 0 0 1 1]
nextCombination = nextComb(A)
the output will be:
A =
1 1 1 1 1 0 1 0 0 1 1
nextCombination =
1 1 1 1 0 1 1 0 0 1 1
You can then use this as a mask into your alphabet (or whatever elements you want combinations of).
C = ['a' 'b' 'c' 'd' 'e' 'f' 'g' 'h' 'i' 'j' 'k']
C(find(nextCombination))
ans = abcdegjk
The first combination in this ordering is
1 1 1 1 1 1 1 1 0 0 0
and the last is
0 0 0 1 1 1 1 1 1 1 1
To generate the first combination programatically,
n = 11; k = 8;
nextCombination = zeros(1,n);
nextCombination(1:k) = 1;
Now you can iterate through the combinations (or however many you're willing to wait for):
for c = 2:nchoosek(n,k) %// start from 2; we already have 1
nextCombination = nextComb(A);
%// do something with the combination...
end
For your example above:
nextCombination = [1 1 0];
C(find(nextCombination))
for c = 2:nchoosek(3,2)
nextCombination = nextComb(nextCombination);
C(find(nextCombination))
end
ans = ab
ans = ac
ans = bc
Note: I've updated the code; I had forgotten to include the line to move all of the 1's that occur prior to the swapped digits to the beginning of the array. The current code (in addition to being corrected above) is on ideone here. Output for 4 choose 2 is:
allCombs =
1 2
1 3
2 3
1 4
2 4
3 4
I need an algorithm in Matlab which counts how many adjacent and non-overlapping (1,1) I have in each row of a matrix A mx(n*2) without using loops. E.g.
A=[1 1 1 0 1 1 0 0 0 1; 1 0 1 1 1 1 0 0 1 1] %m=2, n=5
Then I want
B=[2;3] %mx1
Specific case
Assuming A to have ones and zeros only, this could be one way -
B = sum(reshape(sum(reshape(A',2,[]))==2,size(A,2)/2,[]))
General case
If you are looking for a general approach that must work for all integers and a case where you can specify the pattern of numbers, you may use this -
patt = [0 1] %%// pattern to be found out
B = sum(reshape(ismember(reshape(A',2,[])',patt,'rows'),[],2))
Output
With patt = [1 1], B = [2 3]
With patt = [0 1], B = [1 0]
you can use transpose then reshape so each consecutive values will now be in a row, then compare the top and bottom row (boolean compare or compare the sum of each row to 2), then sum the result of the comparison and reshape the result to your liking.
in code, it would look like:
A=[1 1 1 0 1 1 0 0 0 1; 1 0 1 1 1 1 0 0 1 1] ;
m = size(A,1) ;
n = size(A,2)/2 ;
Atemp = reshape(A.' , 2 , [] , m ) ;
B = squeeze(sum(sum(Atemp)==2))
You could pack everything in one line of code if you want, but several lines is usually easier for comprehension. For clarity, the Atemp matrix looks like that:
Atemp(:,:,1) =
1 1 1 0 0
1 0 1 0 1
Atemp(:,:,2) =
1 1 1 0 1
0 1 1 0 1
You'll notice that each row of the original A matrix has been broken down in 2 rows element-wise. The second line will simply compare the sum of each row with 2, then sum the valid result of the comparisons.
The squeeze command is only to remove the singleton dimensions not necessary anymore.
you can use imresize , for example
imresize(A,[size(A,1),size(A,2)/2])>0.8
ans =
1 0 1 0 0
0 1 1 0 1
this places 1 where you have [1 1] pairs... then you can just use sum
For any pair type [x y] you can :
x=0; y=1;
R(size(A,1),size(A,2)/2)=0; % prealocarting memory
for n=1:size(A,1)
b=[A(n,1:2:end)' A(n,2:2:end)']
try
R(n,find(b(:,1)==x & b(:,2)==y))=1;
end
end
R =
0 0 0 0 1
0 0 0 0 0
With diff (to detect start and end of each run of ones) and accumarray (to group runs of the same row; each run contributes half its length rounded down):
B = diff([zeros(1,size(A,1)); A.'; zeros(1,size(A,1))]); %'// columnwise is easier
[is js] = find(B==1); %// rows and columns of starts of runs of ones
[ie je] = find(B==-1); %// rows and columns of ends of runs of ones
result = accumarray(js, floor((ie-is)/2)); %// sum values for each row of A
I have a 3D image, divided into contiguous regions where each voxel has the same value. The value assigned to this region is unique to the region and serves as a label. The example image below describes the 2D case:
1 1 1 1 2 2 2
1 1 1 2 2 2 3
Im = 1 4 1 2 2 3 3
4 4 4 4 3 3 3
4 4 4 4 3 3 3
I want to create a graph describing adjaciency between these regions. In the above case, this would be:
0 1 0 1
A = 1 0 1 1
0 1 0 1
1 1 1 0
I'm looking for a speedy solution to do this for large 3D images in MATLAB. I came up with a solution that iterates over all regions, which takes 0.05s per iteration - unfortunately, this will take over half an hour for an image with 32'000 regions. Does anybody now a more elegant way of doing this? I'm posting the current algorithm below:
labels = unique(Im); % assuming labels go continuously from 1 to N
A = zeros(labels);
for ii=labels
% border mask to find neighbourhood
dil = imdilate( Im==ii, ones(3,3,3) );
border = dil - (Im==ii);
neighLabels = unique( Im(border>0) );
A(ii,neighLabels) = 1;
end
imdilate is the bottleneck I would like to avoid.
Thank you for your help!
I came up with a solution which is a combination of Divakar's and teng's answers, as well as my own modifications and I generalised it to the 2D or 3D case.
To make it more efficient, I should probably pre-allocate the r and c, but in the meantime, this is the runtime:
For a 3D image of dimension 117x159x126 and 32000 separate regions: 0.79s
For the above 2D example: 0.004671s with this solution, 0.002136s with Divakar's solution, 0.03995s with teng's solution.
I haven't tried extending the winner (Divakar) to the 3D case, though!
noDims = length(size(Im));
validim = ones(size(Im))>0;
labels = unique(Im);
if noDims == 3
Im = padarray(Im,[1 1 1],'replicate', 'post');
shifts = {[-1 0 0] [0 -1 0] [0 0 -1]};
elseif noDims == 2
Im = padarray(Im,[1 1],'replicate', 'post');
shifts = {[-1 0] [0 -1]};
end
% get value of the neighbors for each pixel
% by shifting the image in each direction
r=[]; c=[];
for i = 1:numel(shifts)
tmp = circshift(Im,shifts{i});
r = [r ; Im(validim)];
c = [c ; tmp(validim)];
end
A = sparse(r,c,ones(size(r)), numel(labels), numel(labels) );
% make symmetric, delete diagonal
A = (A+A')>0;
A(1:size(A,1)+1:end)=0;
Thanks for the help!
Try this out -
Im = padarray(Im,[1 1],'replicate');
labels = unique(Im);
box1 = [-size(Im,1)-1 -size(Im,1) -size(Im,1)+1 -1 1 size(Im,1)-1 size(Im,1) size(Im,1)+1];
mat1 = NaN(numel(labels),numel(labels));
for k2=1:numel(labels)
a1 = find(Im==k2);
for k1=1:numel(labels)
a2 = find(Im==k1);
t1 = bsxfun(#plus,a1,box1);
t2 = bsxfun(#eq,t1,permute(a2,[3 2 1]));
mat1(k2,k1) = any(t2(:));
end
end
mat1(1:size(mat1,1)+1:end)=0;
If it works for you, share with us the runtimes as comparison? Would love to see if the coffee brews any faster than half an hour!
Below is my attempt.
Im = [1 1 1 1 2 2 2;
1 1 1 2 2 2 3;
1 4 1 2 2 3 3;
4 4 4 4 3 3 3;
4 4 4 4 3 3 3];
% mark the borders
validim = zeros(size(Im));
validim(2:end-1,2:end-1) = 1;
% get value of the 4-neighbors for each pixel
% by shifting the images 4 times in each direction
numNeighbors = 4;
adj = zeros([prod(size(Im)),numNeighbors]);
shifts = {[0 1] [0 -1] [1 0] [-1 0]};
for i = 1:numNeighbors
tmp = circshift(Im,shifts{i});
tmp(validim == 0) = nan;
adj(:,i) = tmp(:);
end
% mark neighbors where it does not eq Im
imDuplicates = repmat(Im(:),[1 numNeighbors]);
nonequals = adj ~= imDuplicates;
% neglect the border
nonequals(isnan(adj)) = 0;
% get these neighbor values and the corresponding Im value
compared = [imDuplicates(nonequals == 1) adj(nonequals == 1)];
% construct your 'A' % possibly could be more optimized here.
labels = unique(Im);
A = zeros(numel(labels));
for i = 1:size(compared,1)
A(compared(i,1),compared(i,2)) = 1;
end
#Lisa
Yours reasoning is elegant, though it obviously gives wrong answers for labels on the edges.
Try this simple label matrix:
Im =
1 2 2
3 3 3
3 4 4
The resulting adjacency matrix , according to your code is:
A =
0 1 1 0
1 0 1 1
1 1 0 1
0 1 1 0
which claims an adjacency between labels "2" and "4": obviously wrong. This happens simply because you are reading padded Im labels based on "validim" indices, which now doesn't match the new Im and goes all the way down to the lower borders.
There are two vectors:
a = 1:5;
b = 1:2;
in order to find all combinations of these two vectors, I am using the following piece of code:
[A,B] = meshgrid(a,b);
C = cat(2,A',B');
D = reshape(C,[],2);
the result includes all the combinations:
D =
1 1
2 1
3 1
4 1
5 1
1 2
2 2
3 2
4 2
5 2
now the questions:
1- I want to decrease the number of operations to improve the performance for vectors with bigger size. Is there any single function in MATLAB that is doing this?
2- In the case that the number of vectors is more than 2, the meshgrid function cannot be used and has to be replaced with for loops. What is a better solution?
For greater than 2 dimensions, use ndgrid:
>> a = 1:2; b = 1:3; c = 1:2;
>> [A,B,C] = ndgrid(a,b,c);
>> D = [A(:) B(:) C(:)]
D =
1 1 1
2 1 1
1 2 1
2 2 1
1 3 1
2 3 1
1 1 2
2 1 2
1 2 2
2 2 2
1 3 2
2 3 2
Note that ndgrid expects (rows,cols,...) rather than (x,y).
This can be generalized to N dimensions (see here and here):
params = {a,b,c};
vecs = cell(numel(params),1);
[vecs{:}] = ndgrid(params{:});
D = reshape(cat(numel(vecs)+1,vecs{:}),[],numel(vecs));
Also, as described in Robert P.'s answer and here too, kron can also be useful for replicating values (indexes) in this way.
If you have the neural network toolbox, also have a look at combvec, as demonstrated here.
One way would be to combine repmat and the Kronecker tensor product like this:
[repmat(a,size(b)); kron(b,ones(size(a)))]'
ans =
1 1
2 1
3 1
4 1
5 1
1 2
2 2
3 2
4 2
5 2
This can be scaled to more dimensions this way:
a = 1:3;
b = 1:3;
c = 1:3;
x = [repmat(a,1,numel(b)*numel(c)); ...
repmat(kron(b,ones(1,numel(a))),1,numel(c)); ...
kron(c,ones(1,numel(a)*numel(b)))]'
There is a logic! First: simply repeat the first vector. Secondly: Use the tensor product with the dimension of the first vector and repeat it. Third: Use the tensor product with the dimension of (first x second) and repeat (in this case there is not fourth, so no repeat.