Build array from descriptive data without a loop

2 Ansichten (letzte 30 Tage)
Gabriel Stanley
Gabriel Stanley am 17 Feb. 2023
Kommentiert: Walter Roberson am 17 Feb. 2023
I want to go from
Array1 = [10,3,3;1000,178,4];
to
Array2 = [10;13;16;1000;1178;1356;1534];
without using
Idx2 = 1;
Array2 = zeros(sum(Array1(:,3)),1);
for Idx1 = 1:size(Array1,1)
Array2(Idx2:Idx2+Array1(Idx1,3)-1) = [Array1(Idx1,1),Array1(Idx1,1)+[1:Array1(Idx1,3)-1].*Array1(Idx1,2)];
Idx2 = Idx2+Array1(Idx1,3)-1;
end
Help?

Akzeptierte Antwort

Walter Roberson
Walter Roberson am 17 Feb. 2023
Array1 = [10,3,3;1000,178,4];
Array2 = cell2mat(arrayfun(@(Idx) Array1(Idx,1) + (0:Array1(Idx,3)-1).'*Array1(Idx,2), (1:size(Array1,1)).','uniform', 0))
Array2 = 7×1
10 13 16 1000 1178 1356 1534
  2 Kommentare
Gabriel Stanley
Gabriel Stanley am 17 Feb. 2023
Kinda disappointed this hasn't ended up being significantly faster in my use-case, but thank you for providing an answer. I never would've thought to use arrayfun (or cellfun & the like) because I've gotten the impression they're generally slower than a loop.
Walter Roberson
Walter Roberson am 17 Feb. 2023
You didn't ask for performance, you asked for not using a loop. For most operations (but not all, not if you know the right obscure forms), arrayfun and cellfun are slower than looping.
If you were looking for performance, then your existing code could be tweeked to take advantage of cumsum() instead of calculating the indices each time, and you could use the calculation I used instead of using [original, colon expression] list constructor.

Melden Sie sich an, um zu kommentieren.

Weitere Antworten (1)

Kevin Holly
Kevin Holly am 17 Feb. 2023
Array1 = [10,3,3;1000,178,4];
Array2 = cumsum(Array1,2)
Array2 = 2×3
10 13 16 1000 1178 1182
  2 Kommentare
Walter Roberson
Walter Roberson am 17 Feb. 2023
Tthe third column is the number of elements to generate, with the difference being the second column, and the starting point being the first column.
Kevin Holly
Kevin Holly am 17 Feb. 2023
ah, thanks for pointing that out.

Melden Sie sich an, um zu kommentieren.

Kategorien

Mehr zu Matrix Indexing finden Sie in Help Center und File Exchange

Produkte


Version

R2019b

Community Treasure Hunt

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

Start Hunting!

Translated by