Best way to append regression results

Hi there!

I am running a linear regression in a finite loop.

Each time, it produces coefficient estimates and 95% confidence interval ranges.

I would like to append these regression results on a separate matrix.

What is the best way to do this?

For example, suppose I run:

%%%
for i = 1:3
    N = i; % this is the polynomial degree
    [p,S] = polyfit(x, y, N);
%%%

I would like to store the coefficients stored in p.

For the 3rd-degree, I have 1x4 double as class for p.

So what is the best way to append 1x4 double class variable into a clean matrix?

The main issue I am having is the dimensions not matching, something along the line with...

"Unable to perform assignment because

the size of the left side is 1-by-1

and the size of the right side is

1-by-2."

Please help...!! Thank you!

Answer 1

[KSSV]

P = cell(3,1) ; 
for i = 1:3
    N = i; % this is the polynomial degree
    [p,S] = polyfit(x, y, N);
    P{i} = p ; 
%%%

[Tatte Berklee]

KSSV, thanks for a simple answer. But, I actually need to export these results into Excel. So ideally, I would like to have each row to contain the coefficients where the row represents the i-th iteration in the for loop. I tried 'cell2mat(P)', but then it obviously the coefficients all mashed up in one long row. If I am thinking of exporting neatly into Excel, what might be the best way? I prefer each iteration's results stored on a row, the next iteration regression coefficient estimates on the following row, etc.

[Tatte Berklee]

Hi KSSV, on a related note, if I want to store 2 by 2 double class (e.g. confidence interval), then how would I go about storing this information on a clean matrix? Is it storable on cell as you suggested? It doesn't seem to work.

[KSSV]

You can write P which is a cell into excel using xlswrite/ writetable.