How to efficiently replace value at the n last column?
1 visualizzazione (ultimi 30 giorni)
Mostra commenti meno recenti
balandong
il 4 Set 2018
Commentato: balandong
il 4 Set 2018
Dear all, The objective was to replace the value at the end of a column. I have make a simple case but wonder if there is another way to make it more compact and efficient?
Thanks in advance
f_sbj=2
sA=23;
A =1:sA;
appndOut=cell(sA,1);
for f_x=1:sA
Out = nchoosek(A, f_x);
[rows, columns] = find(Out==f_sbj);
Out(rows, :) = [];
[m,n] = size(Out );
newNan=nan(m,(sA-n));
appndOut{f_x}=[newNan Out];
end
2 Commenti
Walter Roberson
il 4 Set 2018
That code appears to insert at the beginning of columns, not replace at the end of columns.
The code can be made more compact, but not necessarily more efficient.
Risposta accettata
Walter Roberson
il 4 Set 2018
lastNcol = @(M, N) M(:,end-N+1:end);
NanPrePad = @(M, N) lastNcol( [nan(size(M,1), N), M], N);
select_and_prepad = @(M, N) NanPrePad( M(all(M~=f_sbj, 2),:) );
appndOut = arrayfun(@(f_x) select_and_prepad(nchoosek(A, f_x), sA), (1:sA).', 'uniform', 0);
You can use shorter names for the anonymous functions if you find the above to not be compact enough.
Though I wonder why you do not do something like
A = setdiff(1:sA, f_sbj);
That would eliminate the possibility of f_sbj being generated by the nchoosek, so it would not be necessary to delete any rows containing f_sbj.
3 Commenti
Walter Roberson
il 4 Set 2018
f_sbj=2
sA=23;
A =1:sA;
lastNcol = @(M, N) M(:,end-N+1:end);
NanPrePad = @(M, N) lastNcol( [nan(size(M,1), N), M], N);
select_and_prepad = @(M, N) NanPrePad( M(all(M~=f_sbj, 2),:), N );
appndOut = arrayfun(@(f_x) select_and_prepad(nchoosek(A, f_x), sA), (1:sA).', 'uniform', 0);
This takes a moment to execute.
Più risposte (0)
Vedere anche
Categorie
Scopri di più su Get Started with MATLAB in Help Center e File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!