Question

我有一个大数组，行数会有所不同，我想拆分并导出到多个文件。

我正在考虑使用reshape命令，但后来我意识到这需要具有相同行数的数组，但并非总是如此。

clear all, clc,clf,tic
num_elm = 11;
num_elm_split = 4;  %Splits into columns
t = linspace(1, num_elm, num_elm)';

v = reshape(t, num_elm_split,[]); %Will split array into different columns
%'for' loop to split number of elements
for ii = 1:length(t(:, 1))
    ii
end

示例：

如果我有一个包含11个值的数组

a = [1; 2; 3; 4; 5; 6; 7; 8; 9; 10; 11]

我希望将它分成每三个值并将数据输出到文件中。

ouput1.txt 会有的 1 2 3

ouput2.txt 会有的 4 五 6

ouput3.txt 会有的 7 8 9

ouput4.txt 会有的 10 11

我知道我可以在Linux中使用split命令，但我只想使用MATLAB / Octave代码。

Answer 1

基本思想是这样做：

values = 1 : 11;
binSize = 3;
fileNum = 1;
n = numel(values);
for i = 1 : binSize : n
   part = values(i : min(n, i + binSize - 1));
   fprintf('File %d contains %s\n', fileNum, mat2str(part));
   fileNum = fileNum + 1;
end

输出：

File 1 contains [1 2 3]
File 2 contains [4 5 6]
File 3 contains [7 8 9]
File 4 contains [10 11]

当然，您需要调整循环中的代码以实际将数据保存在磁盘上。

Answer 2

这是代码......这可能有助于其他人

clear all, clc,clf,tic
filesoxplaylistStr='1soxplaylist.pls'; %playlistist filename fix
dirplstmp='/tmp/tmp/';

values = 1 : 11;
binSize = 3;
fileNum = 0;
n = numel(values);
for ii = 1 : binSize : n
    part = values(ii : min(n, ii + binSize - 1))';
    fileNum = fileNum + 1
    %open file to write to
    fidsoxpl = fopen(strcat(dirplstmp,filesoxplaylistStr), 'w'); %create and open file to write to for sox playlist join file
    for jj=1:length(part)
        part_val=part(jj,1) %gets individual filename or values

        %create sox file to join
        fprintf(fidsoxpl,'File%s=%s%s.wav',num2str(jj),dirplstmp,num2str(part_val,'%06d'));%create playlist data file
        fprintf(fidsoxpl,'\n');

    end
    % close file to write to
    fclose(fidsoxpl);

    fn=strcat('test',num2str(fileNum,'%02d'));
    %join files with sox
    syscmd=strcat({'sox '},dirplstmp ,filesoxplaylistStr, {' '},dirplstmp,fn,{'.wav'});
    system(syscmd);

    %add freq to file name


end

在MATLAB / Octave中拆分数组和导出值

2 个答案: