- Your structures MRET_files(i).name and RSK_files(i).name are not actually different for different values of i, or
- Your function Factor_Returns_Function doesn't actually process the inputs the way you think
How can I effeciently store about 200 months of multivariate regression data?
1 次查看(过去 30 天)
显示 更早的评论
So I am running a regression of monthly returns for about 30000 stocks (each regression is run separately for stocks grouped by country) over anout 9 different factors. I am using this for loop to loop through every file, because each month's data is stored in a differentfile. I am trying to use a structure with fields to represent all of the outputs of my multivariate regression function, but when the regression is complete, every part of the 1x196 structure labeled 'regression_structure' is exactly the same! It just copied over the first month's rgeression 196 times. Where did I go wrong!?
for i=1:num_months
% need to pass file names for monthly data files
file1 = MRET_files(i).name;
file2 = RSK_files(i).name;
[industry_rep_cell, country_rep_array,u_country_codes, u_industries, beta, mv] = Factor_Returns_Function(file1, file2);
regression_structure(i).rep_industries = industry_rep_cell;
regression_structure(i).rep_countries = country_rep_array;
regression_structure(i).all_industries = u_industries;
regression_structure(i).all_countries = u_country_codes;
regression_structure(i).beta_simple = beta;
regression_structure(i).beta_multi = mv.beta1;
end
0 个评论
回答(1 个)
the cyclist
2013-7-2
编辑:the cyclist
2013-7-2
This simple code gives different values for each element of the structure:
for i=1:196
[B,bint] = regress(rand(3,1),rand(3,1));
regression_structure(i).B = B;
regression_structure(i).bint = bint;
end
This test suggests to me that the problem is either
I can't test those hypotheses.
另请参阅
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!