Preallocating and filling long array with shorter arrays

Hello dear community,

I have a question:

If I want to fll an array with single values using preallocation, I do:

input = zeros(1,1000000);
for k = 2:1000000
    input(k) = input(k-1) + 5; % 5 is a newdata value in this example
end 

so, in input will be stored values 0, 5, 10, 15, 20....

*********************************************************************

now, I have a data input from arrays (ex.) 5 values long, and I want to concatenate them to one big input array after every loop:

newdata:

4 5 7 8 9, 5 8 7 9 5, 1 7 8 9 6, 1 2 3 6 5

to one array input in few steps:

4 5 7 8 9 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

............

4 5 7 8 9 5 8 7 9 5 1 7 8 9 6 1 2 3 6 5

If I create a big zeroarray, and then create an input variable in a loop, I just create every time a new variable in a loop. How can I use the method from single variable by arrays to use preallocation?

input = [zeros(20,[])]; %it has no sence 
for i = 1 : total_frames
    input = [input;newdata]; 
end

Thank you!

P.S. Actually, I'm working with big arrays (48000 * 100000000), the short example is just for better understanding. I want to make my algorithm time efficient.

Answer 1

[Stephen23]

The simple apparoach is to use a matrix or array, for example:

N = 10;
M = nan(5,N);
for k = 1:N
    M(:,k) = rand(5,1);
end
display(M)

M = 5×10

0.5932 0.1375 0.9838 0.9635 0.7082 0.7918 0.3489 0.1323 0.9478 0.5962 0.6731 0.8178 0.3265 0.6184 0.2314 0.0253 0.0699 0.5711 0.7934 0.5818 0.8518 0.1121 0.2469 0.3151 0.5395 0.7872 0.9489 0.0019 0.1219 0.3131 0.3972 0.2027 0.3039 0.0395 0.6487 0.6918 0.8420 0.1117 0.2279 0.0301 0.5613 0.1273 0.4150 0.2947 0.1517 0.5669 0.3289 0.1059 0.8498 0.1112

"Actually, I'm working with big arrays (48000 * 100000000)"

https://www.mathworks.com/help/matlab/import_export/tall-arrays.html

[Nik Rocky]

Great answer, thank you very much!

Here is comparsion old to new:

spf = 48000;
frames = 200;                                                                  
data=[nan(spf*frames,[])];                                                   
tic
for k = 1 : frames
    input = rand(48000,1);
    data = [data;input];                                       
end
toc

Elapsed time is 3.206472 seconds.

spf = 48000;
frames = 200;
data = nan(spf,frames);
tic
for k = 1:frames
    data(:,k) = rand(spf,1);
end
data = reshape(data.', [],1);
toc

Elapsed time is 0.093135 seconds.

In the end I have a big (spf*frames) x 1 array, that I need

Feel the difference =)