Hi all,
I converted a nested for loo into a unique parfor loop. The nested for loop looked like this:
for RIPETIZIONE = 1:N_RIPETIZIONI
for k=1:size(lambda_tol_vector)
Now, the parfor loop takes on the following values:
parfor n=1:N_RIPETIZIONI*K
hence merging into a single loop the nested loops of before.
Now, inside the "outer" nested loop (for RIPETIZIONE = 1:N_RIPETIZIONI), but outside the inner loop (for k=1:size(lambda_tol_vector)), there was this part:
scv=size(columns_validation_test,2);
for counter_columns = 1:scv
COLONNA_SELEZIONATA=columns_validation_test(counter_columns);
Ctest=zeros(size(A.Value));
Ctest(:,COLONNA_SELEZIONATA)=C(:,COLONNA_SELEZIONATA);
for j=1:size(lambdavector)
RMSE_initial_test{counter_columns}(j)=sqrt(sum2(Diff_sq_initial{j}.*Ctest)/sum(Ctest(:)));
RMSE_final_corrected_validation{counter_columns}(j)=sqrt(sum2(Diff_sq_corrected{j}.*Cvalidation)/sum(Cvalidation(:)));
RMSE_final_test{counter_columns}(j)=sqrt(sum2(Diff_sq{j}.*Ctest)/sum(Ctest(:)));
RMSE_final_corrected_test{counter_columns}(j)=sqrt(sum2(Diff_sq_corrected{j}.*Ctest)/sum(Ctest(:)));
[~,arg_min_temp_1]=min(RMSE_final_corrected_validation{counter_columns});
which I need now to includde in the unique parfor loop with a pope indexing. Fo instance, Diff_sq_initial, Diff_sq_corrected and Diff_sq are all defined withing the parfor loop with an index of n in the new loop (whereas previosuly they were indexed with k inside the for k=1:size(lambda_tol_vector) loop).
Is there a way to properly indexing those variables so that they fit the new parfor loop?
Thaank you
