hello everyone please can someone help me with stock price prediction. That is I wrote this code and from the I vector , I want a code that will remove indexes from the I vector more than 60 seconds. Thanks

 
LSE_matrix =log(nstock_val); %log of the data
I=1:(size(LSE_matrix,1)-1); % selecting the indices of all prices but the last time  when stock was opened

dLSE_col1 = LSE_matrix(I+1,1) - LSE_matrix(I,1);%  log difference

[Geoff Hayes]

Afua - how or where is time used in the above code? I see that you have valus for stock prices (?) but where is a time array?

[Afua Amoako Dadey]

Thanks for the feedback. This is a section of the code and i wanted to know how i can write a code to remove the time stocks index that gobeyond 60 seconds. That is we remove the indices from the "I" variable seen above. Thanks

[Geoff Hayes]

But what variable describes the time? How, from the above code, can we determine which stock index goes beyond 60 seconds?

[Afua Amoako Dadey]

Atttached is of picture the corresponding times. So say we consider 498 and 498. its more than 60 sec. how do i write a code to remove such an index. Thanks for the help.

[jonas]

Dont you think it would be easier if you upload a sample of the data?

Answer 1

[jonas]

You can adapt this to your needs

A = readtable('LSE1.csv')
t = datetime(A{:,1},'inputformat','dd.MM.yyyy HH:mm:ss.SSS')
data = A{:,2:end};
id = [false;diff(t)>seconds(60)];
data(id,:) = [];
t(id) = [];

If the period between t(n) and t(n+1) is longer than 60s, then the data recorded at t(n+1) is deleted.

[Afua Amoako Dadey]

Hello thanks for the help. I need one last help. How do i get this code to remove from it indices > 60 seconds. For the open, close, high and low simultaneously.

[jonas]

I think that is exactly what it does.

the line

id = [false;diff(t)>seconds(60)];

finds all row indices where the previous row differs by more than 60 seconds, whereas the line

data(id,:) = [];

deletes all data in those rows.

then you are left with a vector of time, t, and a corresponding matrix of data, data.

[Afua Amoako Dadey]

Thanks ... but can i show you one more thing.

%Close prices in matrix 

nstock_val=[Close(1:30928,1) Close1(1:30928,1) Close2(1:30928,1) Close3(1:30928,1) Close4(1:30928,1)]; % reading the data 
t=[Gmttime(1:30928,1) Gmttime1(1:30928,1) Gmttime2(1:30928,1)  Gmttime3(1:30928,1) Gmttime4(1:30928,1)]; % reading the data 
time = datetime(t,'inputformat','dd.MM.yyyy HH:mm:ss.SSS');

LSE_matrix =log(nstock_val); %log of the datanstock_val=[Close(1:30928,1) Close1(1:30928,1) Close2(1:30928,1) Close3(1:30928,1) Close4(1:30928,1)]; % reading the data 
I=1:(size(LSE_matrix,1)-1); % selecting the indices of all prices but the last time  when stock was opened


 data = t(:,1:end);
 id = [false;diff(time(:,1))> seconds(60)]
 data(id,:) = [];

dLSE_col1 = LSE_matrix(I+1,1) - LSE_matrix(I,1);%  log difference

dLSE_col2 = LSE_matrix(I+1,2) - LSE_matrix(I,2);

dLSE_col3 = LSE_matrix(I+1,3) - LSE_matrix(I,3);

dLSE_col4 = LSE_matrix(I+1,4) - LSE_matrix(I,4);

dLSE_col5 = LSE_matrix(I+1,5) - LSE_matrix(I,5);
matrix_logdiff=[dLSE_col1 dLSE_col2 dLSE_col3 dLSE_col4 dLSE_col5];

nsample = floor((2/3)* length(matrix_logdiff)); %two thirds of the log 
% difference of the data
 
matrix_logdiff_nsample = matrix_logdiff(1:nsample,:);

mean_vec = mean(matrix_logdiff_nsample); %finding the mean.

mean_vec = mean_vec';
 
cov_matrix = cov(matrix_logdiff_nsample); %the covariance
dt=1;

%GBM parameter estimate
capbhat = sqrtm(cov_matrix./dt);   

ahat_vec  = (mean_vec/dt)+ diag((capbhat).^2)/2;

s0_vector =[nstock_val(nsample,1) nstock_val(nsample,2) nstock_val(nsample,3) nstock_val(nsample,4) nstock_val(nsample,5)];

s0_vector = s0_vector';

[Afua Amoako Dadey]

where the different close prices are from the different LSEs

[jonas]

I don't understand exactly what the problem is, but I'm not a big fan of how you structure the data. I would put all my .csv files in a separate folder and read them as follows:

addpath('C:\files\') % change this to the correct path
files = dir('C:\files\*.csv'); % again..
for i = 1:numel(files)
    opts = detectImportOptions(files(i).name);
    opts.VariableTypes{1} = 'datetime';
    opts = setvaropts(opts,'GmtTime','inputformat','dd.MM.yyyy HH:mm:ss.SSS');
    A{i} = readtimetable(files(i).name,opts);
end

You will end up with 5 tables stored in cell array A. You can then append them to a single table or loop over each tables, no need to work with "Var1, Var2, Var3..."

[Afua Amoako Dadey]

Okay, Mr Jonas. I really appreciate the help

Answer 2

[Afua Amoako Dadey]

Please find attached