How do I stop accumarray adding rows for non-existent values of the index column?

I want to sum the values in one column of an array based on values in another (index) column. What I get is a row for all values up to the maximum in the index column. For example,

a = [2 1.1; 2 1.2; 4 1.3; 4 1.4]

b = accumarray(a(:,1),a(:,2))

doesn't give what I expect: b =

    2.3000     % i.e. sum of values for which the index is 2
    2.7000     % i.e. sum of values for which the index is 4

but

b =

         0
    2.3000
         0
    2.7000

How do I suppress the rows where the index value doesn't exist? This is a trivial example, but in what I'm trying to do the index values are in the range 195001 to 202012 so I get a large number of zeros. All help gratefully received...

Answer 1

[Steven Lord]

unique the first column and use the third output in accumarray instead of the first column of a.

a = [2 1.1; 2 1.2; 4 1.3; 4 1.4];
b = accumarray(a(:,1),a(:,2))

[c1, ~, c3] = unique(a(:, 1));
d = accumarray(c3, a(:, 2))
results = table(c1, d, 'VariableNames', {'elementInAcolumn1', 'accumulatedValue'})

I assembled the results into a table to more easily show the relationship between c1 and d.

[Ed Rowe]

Great answer, solved my problem and I understand unique and table a bit better now too! Many thanks.

Answer 2

[Jan]

a = [2 1.1; 2 1.2; 4 1.3; 4 1.4];
[~, ~, group] = unique(a(:, 1));
b = accumarray(group, a(:,2))