You can use histcounts or histogram for this purpose. Let's generate some sample data, 100 integers between 1 and 10.
The edges should start at each unique value in A. We need to add one additional edge at the end, otherwise the last bin will count the two largest unique values contained in A, not just the largest. All bins but the last contain their left edge but not their right edge, while the last bin contains both its edges.
U = unique(A);
edges = [U, max(U)+1];
Now let's count. I'll put the results in a table array for easy reading.
counts = histcounts(A, edges);
results = table(U(:), counts(:), 'VariableNames', {'UniqueElements', 'Count'})
If you want to see the results graphically, use histogram instead of histcounts.
h = histogram(A, edges);
counts2 = h.Values;
isequal(counts2, counts)
If you want the bins to be centered on integer values (useful if your data is all integer values, as in the case in the original question) you can specify the 'integers' BinMethod instead of building the edges yourself. If you do this with histcounts you'd want to call it with two outputs so you receive both the bin counts and the vector of bin edges. Remember that the vector of bin edges will be one element longer than the vector of bin counts since bin edges are fence posts and bin counts are fence rails. h = histogram(A, 'BinMethod', 'integers');
[counts3, edges3] = histcounts(A, 'BinMethod', 'integers')
