You made a common novice user mistake: mistaking (row, column) for (x, y). Note: row is Y, not X. and column is X, not Y. ind2sub() returns [row, column], NOT [X, Y]. So you might be tempted to do:
[Y, X] = ind2sub(size(c), max_idx)
however even that is not right. Why not? Because max_idx is not the linear index over the whole matrix - it is the column number of each row. For example while max() might give [4,4,1], the linear indexes would be [4,8,9]. But that is not what you were passing in to ind2sub, you were passing in [4,4,1], so it didn't give you the expected output.
To correct the code, see the solution Star Strider gave you.
