Question

Not sure how to do this and it’s really driving me crazy.
I have a large matrix with 3 columns showing:

1. The number of items sold in the year
2. The first year the items were sold
3. The last year the items were sold.

So in the following example, 13 items were sold each year between 2000-2010 and 44 in 2003 and 2004:

Matrix_A = {13,    2000,    2010;
            44,    2003,    2004}

I’m trying to separate out the values by year to calculate the total number of items sold each year. So, 57 (13 + 44) items will have been sold in 2003 and 2004, and 13 all other years.

The resulting matrix would look like this:

{13, 2000;
 13, 2001;
 13, 2002;
 13, 2003;
 13, 2004;
 13, 2005;
 13, 2006;
 13, 2007;
 13, 2008;
 13, 2009; 
 13, 2010;
 44, 2003;
 44, 2004}

---

I’ve tried creating a separate empty matrix for each year, looping through Matrix_A and assigning the row to the appropriate annual matrix. So, for Matrix_A:

• 13 will be added to the matrixes for 2000 to 2010 and,
• 44 to the matrices for 2003 and 2004.

But this seems to involve dynamic variable names, which I can’t implement.

In all, I’m lost. Any ideas?

Thanks!

Answer 1

Here is a simple loop over the datasets that generates the matrix you showed in the question

B = [];
A = [13,    2000,    2010;
     44,    2003,    2004];

for i=1:size(A,1)
    years = [A(i,2):A(i,3)]';
    start = size(B, 1);
    B(start+1:start+numel(years),2) = years;
    B(start+1:start+numel(years),1) = A(i,1);
end
B =

      13        2000
      13        2001
      13        2002
      13        2003
      13        2004
      13        2005
      13        2006
      13        2007
      13        2008
      13        2009
      13        2010
      44        2003
      44        2004