Dynamically adding element to an array

Dear Community,

Still new to matlab and exploring, It may be a simple problem but I am stuck for sometime on trying to come out with a clean logic .

Say i have an array (N X 2 )

signal price

1 29.21

-1 29.41

-1 29.43

-1 29.63

1 29.01

0 29.21

1 28.62

1 28.82

I wanted to add a row if the consecutive row in column 1 is -1 to 1 or -1 to 1 , the row to be added is a 0 and column 2 will be the price in the next row.

Basically a 1 is a long signal and -1 is a short signal and 0 is exit a position (in finance). The problem with the original data is when the scenario is from long to short ( 1 to -1) , i need to exit it before shorting , so i need to add a row with 0 and the price to exit.

The result I am expecting will be :

signal price

1 29.21

0 29 41 (added row)

-1 29.41

-1 29.43

-1 29.63

0 29.01 (added row)

1 29.01

0 29.21

1 28.62

1 28.82

The code I attempted :

spread = y;   %price
temp=NaN(length(spread), 2); % initialise N X 2 array

temp(:,2)=spread; % put price into 2nd column

%numUnits is a single row of signals (eg : -1,0,1...)

for i = 2:numel(numUnits)
       
     temp(i,1)=numUnits(i); 
    
      compare = abs(numUnits(i)-numUnits(i-1));
    
      if  compare == 2
        
        temp = [temp(1:i-1,:); temp(i:i,:); temp(i:end,:)];  
        temp(i,1)=0;
        
        %temp(i,2)=i;
      end
end

The result is incorrect because the for loop is always operating on the original NumUnits array and not the Temp where I am adding the extra rows. I dont have an elegant solution to add dynamically an array and operate it on the fly.

Anyone able to help or give some clue? Thanks in advance!

Darren

Answer 1

[Timo Dietz]

priceArray= ...
    [1   29.21
    -1  29.41
    -1  29.43
    -1  29.63
    1   29.01
    0   29.21
    1   28.62 
    1   28.82]

y=priceArray(:, 2);
numUnits = priceArray(:, 1);

% YOUR CODE WITH MODIFICATIONS:

spread = y;   %price
temp=NaN(2*length(spread), 2); % initialise N X 2 array TWICE the size of the origin
%temp(:,2)=spread; % put price into 2nd column
%numUnits is a single row of signals (eg : -1,0,1...)

idx = 1;  % introduce second index variable
for i = 2:numel(numUnits)
       
     %temp(i,1)=numUnits(i); 
    
      compare = abs(numUnits(i)-numUnits(i-1))
    
      if  compare == 2
          temp(idx, :) = priceArray(i-1, :);
          idx = idx + 1;
          temp(idx, :) = [0 priceArray(i, 2)];
          
          
      else
          temp(idx, :) = priceArray(i-1, :);
      end
      idx = idx + 1;
end
temp(idx, :) = priceArray(i, :); % add last row which always equals the origin
temp = temp(1:idx, :)

Try these modifications. The point is that you have to introduce a second index variable since your temp array is bigger than

the origin. So you can't use one common index. Further, the initialization of temp must reserve more rows - I chose twice

the size for simplicity (+50% should suffice already).

[Darren Lim]

thank you Timo, this works! on my learning journey and this helps a lot!