Formula Mean Values in Matlab code.

Hello, can anyone help to write in Matlab code the formula below. The third term with the minus sine is a double sum.

(sum from i=1 to N of Xt-i*168)/N + (sum from j=1 to 7 of Xt-j*24)/7 -(sum of i=1 to n sum of j=1 to 7 of Xt-i*168-j*24)/(7*N).

thank you for your time

[Azzi Abdelmalek]

what is Xt? its size?

[aggelos]

hi, X it's a variable and t is the associated index. In the code below X is L.

Dear Azzi, My indexes for the sum's are i and j. t Stands for time . Xt is a time series of hourly spot electricity prices.

[Azzi Abdelmalek]

what are your index i or t?

[aggelos]

I have the code for a similar formula but i can;t understand it . Here it is. The formula is (sum from i=1 to N of Xt-i*7)/N + (sum from j=1 to 7 of Xt-j7)/7 -(sum of i=1 to n sum of j=1 to 7 of Xt-i*7-j)/(7*N). and the code is

mean(L(start+i-7:-7:start+i-N*7)) + ... mean(L(start+i-1:-1:start+i-7)) - mean(L(start+i-7-1:-1:start+i-(N+1)*7));

Answer 1

[Image Analyst]

I don't think you've explained this correctly. What's the difference between t and i? Anyway, here's some code to get you started:

N = 100; % Or whatever.  The number of elements.
T = 1 : N;
Xt = 2 * T;  % Or whatever function of T that X is.
i = 1 : N;
theFirstSum = sum(Xt - (i * 168) / N)

[aggelos]

Thank you very much for tour time. Now I have to go , but i will try to correct it later.

[aggelos]

As I said before Xt is a time series of electricity hurly prices. I need ths formula to break the time seris in two parts . a deterministic seasonal component which comes from this formula and a stochastic component ,which is the difference of the original time series minus the lagged series corresponding to the formula ineed to translate into matlab code.

[Image Analyst]

OK. So, didn't I do that for the first sum? I just had to guess at Xt because you didn't tell us what it was - you replace that line with your actual Xt. Then you just basically do the same for the other sums. After all, it's your assignment so I can't just do everything for you and hand it over to you.

[aggelos]

Dear Image Analyst, To make the things clear. Forget The 't' index.Let's say that X is my time series (L in the code above). My problem is how to define the indexes in each sum. For example, in the first sum the indexes are from i=1 to N , but i need to take values of the form : for i=1 I need to take the price which is 168 hours before. for i =2 the price which is 2*168=336 hours before ... for i=N the price which is N*168 hours before. Sorry to bother you and thank you very much for your time and interest.

[Image Analyst]

To sum the prior 168 items, in general

theSum = zeros(1, lastIndex);
for k = 168 : lastIndex
  theSum(k) = = sum(Xt((k - 167) : k));
end