how to concatenate two tables efficiently

Question

Salma fathi 2022-7-6

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1754365-how-to-concatenate-two-tables-efficiently

回答： Campion Loong 2022-7-21

I am trying to read a huge amount of files, each file would have data stored in a atable in it, after reading the files we would like o put all the tables from the files into one big table, to concatenate the table i am using the follwoing lines. They work just fine but the issue is when the table size get much bigger the loop will be running relatively slow and unefficintely. Is there any more efficient way to do it?

        if ifile==3
            EDPAll=EDPT(:,[1,6:7]);
        else
            EDPAll=[EDPAll;EDPT(:,[1,6:7])];
        end        

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Jonas 2022-7-6

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1754365-how-to-concatenate-two-tables-efficiently#answer_1001235

编辑：Jonas 2022-7-6

在 MATLAB Online 中打开

before concatenation you can go trough all tables and add up the number of rows. Then preallocate the table using three columns and the number of rows you found. Then assign the values explicitly using indexing, e.g.

% -> preallocate here the table in suitable size e.g. EDPAll  <-
currFirstIdx=1;
for nr=1:numOfTables
    % -> here, retrieve table number nr, e.g. EDPT <-
    
    EDPAll(currFirstIdx:currFirstIdx+height(EDPT)-1,:)=EDPT;
    currFirstIdx=currFirstIdx+height(EDPT);
end

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

Answer 2

Campion Loong 2022-7-21

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1754365-how-to-concatenate-two-tables-efficiently#answer_1012565

"I am trying to read a huge amount of files, each file would have data stored in a atable in it"

Have you tried using Datastore to manage this workflow? Use tabularTextDatastore if all your files are text in tabular data format. You can use the "SelectedVariableNames" Name/Value pair to read in only the variables you can about (i.e. rather than read everything and then only select [1,6:7]), then read everything with readall() or read only up to the size (i.e. ReadSize) you are interested in