find sum of all the entries with value 0

I have a cell matrix of mat = [ 1 0 0 4;0 2 0 4], i want to find the sum of the all the entries having value 0. in the case of given example the entries having count 0 are 4; so sum will be 4. how can i do this.

Answer 1

[Matt J]

 sum(~mat(:))

[arun]

sum(sum(~mat))

Thank You.

[Matt J]

No, what I proposed

sum(~mat(:))

works indepedently of ndims(mat) and will save you additional function calls to sum().

[arun]

okieee,sorry............... i got it.

Answer 2

[Jos (10584)]

NNZ is the dedicated function to do this:

nnz(~mat)

[Matt J]

NNZ is optimal for sparse matrices. For full, sum is faster,

A=sprand(1e4,1e4,1e-3)>0;
Af=full(A);

tic;nnz(A);toc
%Elapsed time is 0.000031 seconds.   

tic;sum(A(:));toc
%Elapsed time is 0.000491 seconds.

tic;nnz(Af);toc
Elapsed time is 0.083754 seconds.

tic;sum(Af(:));toc
Elapsed time is 0.053861 seconds.

[John D'Errico]

Note that nnz(~mat) will be poor even for large sparse matrices, because if mat is sparse, then ~mat must create a large and rather dense logical matrix. Better then would be to use a form like:

numel(mat) - nnz(mat)

In my tests, for a large fairly sparse matrix like that generated by Matt in his comment, my form was the fastest.

[Jos (10584)]

Thanks, Matt and John, for the insights.