inefficient loop to vertically concatenate tables
显示 更早的评论
I have a 300000 by 2 table, called output, that looks like in the screenshot below. Each entry in the 2nd column as inside it a 3-column table with different numbers of rows. I woud like to havee one single table with 3-columns, that is, vertically merge all those little 3-column tables into one.
So far, I was doing:
tableBig = array2table([]);
for ii = 1 : size(output,1)
temp = output{ii,2};
temp = temp{:};
tableBig = [tableBig; temp];
end
The problem is that this loop is a inefficient time wise. The first iterations take in the order of 0.0025 seconds which is not too too bad. But as tableBig gets bigger, iterations seem to take longer and longer, so that projected time based on tic/toc is more than 24 hours.
Is there a way to do this without a loop or to speed up the loop?
Thanks
采纳的回答
Jos (10584)
2019-8-8
You can apply comma-separated list expansion to tables too, so this one-liner should work.
tableBig = cat(1, output{:,2})
5 个评论
Guillaume
2019-8-8
or
tableBig = vertcat(output{:, 2})
Jos (10584)
2019-8-8
I always forget about all those fancy newer functions that are wrappers around the basic ones ;-D
Daniel Pinto
2019-8-8
Thanks to both. However, when I do that, I get the following. I would like to have as an output all the contents of the small tables, i.e., a very long table with 3 columns.
I'd like to have the contents of the 6 by 3 matrix below the contents of the 12 by 3 matrix and so on.
Is that feasible?
Guillaume
2019-8-8
Oh, yes of course, the tables are stored inside a cell array in the enclosing table. You'll need another cat|vertcat:
c = vertcat(output{:, 2}); %or cat(1, output{:, 2}); %it's the same
tableBig = vertcat(c{:}); %or cat(1, c{:}); %it's the same
I don't think you can do it efficiently in just one step.
Daniel Pinto
2019-8-8
更多回答（0 个）
类别
在 帮助中心 和 File Exchange 中查找有关 Loops and Conditional Statements 的更多信息
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!
Translated by 
