CODEWITHSUNDEEP
matlab
Bassam
Bassam

Reputation: 1

How can I use a combination of the min and find functions in Matlab to find how many zeroes are in the vector?

Title says it all. I Know what both of these functions do individually, but combining them is proving difficult.

Upvotes: 0

Views: 62

Answers (1)

rayryeng
rayryeng

Reputation: 104514

Collecting all of the answers from the comments, you can do it in a variety of ways:

  1. numel(find(a == min(0))) (à la Try Hard). The min(0) is rather superfluous as you can simply do a == 0, but because you are required to use min, we need that there. The general case is (à la Divakar) numel(find(a == min(abs(a))));. This is assuming that the minimum of a is zero. Essentially, this finds all locations that are zero, and determining how many of these elements are zero is done with numel.

  2. I personally would not use find and min. A.Donda suggests to use sum, which is certainly more elegant... so do: sum(a == 0). a == 0 produces a logical vector where 1 at a position means that the corresponding element at this position in a is 0 and 0 otherwise. We sum up all of the cases where we have 1, which produces how many zero values there are.

  3. This was one I came up with: numel(a) - nnz(a). This comes from the fact that we can decompose the total number of elements in a vector to be the number of zero elements, added with the total of non-zero elements, or nz(a) + nnz(a) = numel(a), and so nz(a) = numel(a) - nnz(a). There is a function in MATLAB called nnz, which counts how many non-zero elements there are in a vector or matrix. Therefore, you can count how many zero elements there are by numel(a) - nnz(a). It's not as elegant as sum(a == 0), but it's certainly another way to achieve what you want.

Upvotes: 1

Related Questions