This function is supposed to output smallestPosition = 1 , 3 when given: smallestPosition = FindSmalle​stElevatio​nChange(3,​[2,7,4,5])​, however I get returned smallestPosition = 1,0,3 instead, cant seem to get rid of the random zero

function [smallestPosition] = FindSmallestElevationChange(currentPosition,FuturePositions)
%This function will calculate the difference between current position, and
%available positions, in reference to elevation, by considering an array of
%values and returning the smallest difference.
% Inputs : currentPosition - single integer
%          FuturePositions - 1D Array
% Outputs : smallestPosition - 1D array with smallest elevation change
%Author: Tai Leong

%calculate difference between currentPostion value and each element of
%FuturePositions
for i = 1: length(FuturePositions)
    if FuturePositions(i) >= currentPosition
        change = FuturePositions(i) - currentPosition;
    else 
        change = currentPosition - FuturePositions(i);
    end
    elevationChange(i) = change;
end

%set smallest difference(s) into 1D array
smallestvalue = min(elevationChange);
 for j = 1 : length(elevationChange)

          if (elevationChange(j) == smallestvalue)%display position that produces smallest elevation change

              smallestPosition(j) = j;
          end 
   end

Answer 1

[Jan]

smallestPosition = [];
for j = 1 : length(elevationChange)
  if (elevationChange(j) == smallestvalue)
     % Not the j.th element:  smallestPosition(j) = j;
     % But append to the output:
     smallestPosition = [smallestPosition, j];
  end 
end

You will see a small red mark in the editor and a message, which tells you that the iterative growing of a vector is not efficient. For such tiny problems this does not matter. But for productive code, this is better:

smallestPosition = zeros(size(elevationChange));
for j = 1 : length(elevationChange)
  if (elevationChange(j) == smallestvalue)
     smallestPosition(j) = j;
  end 
end
smallestPosition = smallestPosition(smallestPosition ~= 0);

Finally note that the abs() command would be much cleaner and faster than the first loop, and that the loops are not needed at all. I assume that this is a homework - this is not the only thread concerning this question. An efficient solution needs 2 lines of code only and it is hard to give any further hints without posting the complete solution.