How to declare a variable that changes it's size after each loop iteration?

1 view (last 30 days)
hoque
hoque on 7 May 2016
Commented: Weird Rando on 11 May 2016
I want to declare/pre-allocate two variables "delay_points_local","CFO_points_local" that change the sizes. How can I declare it? Here is the code:
pdf_tau = step_CFO*sum(g_prime_bar_all_relays(:,:,k),1);
[~,tau_k_pos] = max(pdf_tau);
arg_max_tau(k) = delay_points(1,tau_k_pos);
pdf_CFO = step_tau*sum(g_prime_bar_all_relays(:,:,k),2);
[~,CFO_k_pos] = max(pdf_CFO);
arg_max_CFO(k) = CFO_points(1, CFO_k_pos);
step_CFO = 0.0001;
step_tau = 0.01;
rho_g = 20;
for k = 1 : K
delay_points_local(k,:) = arg_max_tau(k) - 0.3 : step_tau : arg_max_tau(k) + 0.3;
CFO_points_local(k,:) = arg_max_CFO(k) - 0.005 : step_CFO : arg_max_CFO(k) + 0.005;
[g_prime_bar_all_relays_local(:,:,k)] = Importance_function_grid( y_rd , CFO_points_local(k,:) , delay_points_local(k,:) , L , os_factor , K , TrainingSequence(k,:) , rho_g , step_tau , step_CFO,roll_off);
end

Answers (2)

CS Researcher
CS Researcher on 7 May 2016
I am not completely sure I understand your question but you can do it two different ways:
If the number of columns are fixed:
delay_points_local = zeros(K,N);
for k = 1:K
delay_points_local(k,:) = ... (the way you are doing it)
end
If you are not sure of the column size
delay_points_local = cell(1,K);
for k = 1:K
delay_points_local{1,k} = ...
end
Hope this helps!
  3 Comments
hoque
hoque on 8 May 2016
The change is variable. I declared it with max column size. With that max column size, I could generate the C code successfully but the Matlab code shows error showing the message that "Subscripted assignment dimension mismatch". Is that means that the generated C code will work wrong? My work starts from C code.

Sign in to comment.


Weird Rando
Weird Rando on 7 May 2016
Dunno if this code would run. Basically I modify the for loops and added two variable (temp_delay_points_local, temp_CFO_points_local). The delay_points_local and CFO_points_local will contain 0 padding due to the uneven sizes of the array. It may be better to store them in a cell.
pdf_tau = step_CFO*sum(g_prime_bar_all_relays(:,:,k),1);
[~,tau_k_pos] = max(pdf_tau);
arg_max_tau(k) = delay_points(1,tau_k_pos);
pdf_CFO = step_tau*sum(g_prime_bar_all_relays(:,:,k),2);
[~,CFO_k_pos] = max(pdf_CFO);
arg_max_CFO(k) = CFO_points(1, CFO_k_pos);
step_CFO = 0.0001;
step_tau = 0.01;
rho_g = 20;
for k = 1 : K
temp_delay_points_local = arg_max_tau(k) - 0.3 : step_tau : arg_max_tau(k) + 0.3;
delay_points_local(k,1:numel(temp_delay_points_local)) = temp_delay_points_local;
temp_CFO_points_local = arg_max_CFO(k) - 0.005 : step_CFO : arg_max_CFO(k) + 0.005;
CFO_points_local(k,1:numel(temp_CFO_points_local)) = temp_CFO_points_local;
[g_prime_bar_all_relays_local(:,:,k)] = Importance_function_grid( y_rd , CFO_points_local(k,:) , delay_points_local(k,:) , L , os_factor , K , TrainingSequence(k,:) , rho_g , step_tau , step_CFO,roll_off);
end
  2 Comments

Sign in to comment.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by