How to export to csv in multiple rows instead of all columns?

Hello all, I am developing a code to export a long (>32,000 values) comma separated value char variable set of ASCii scope measurements into a comma separated value, which is saved in Excel. Excel limits its # of columns to 16,384, but rows are limited to 1,048,576. For this reason I want to either split the data into multiple rows, or save all data into column A as rows. This is my export code:

    function csvExportData(channel,csvData)
        if (channel == 1)
            splitData = strsplit(csvData,','); %remove commas
            exportData = mat2dataset(splitData); %creates dataset array

            export(exportData,'file','scopeDataTestCh1.csv','Delimiter',',','WriteVarNames',false); %write to scopeDataTest.csv 
        end
    end

csvData is in the form: 2.79E-02,1.63E-02,5.8E-03,-2.6E-03,-1.11E-02............... where the values vary, and the length varies, sometimes being very long. I can not have a limit to the length of the data, as its important that I can get all the y-values for the scope's measurement (up to 80 GSa/s).

Thanks in advance for any tips y'all can provide.

[John BG]

chop the long 1D data string into reasonable time base unit: miliseconds, seconds, whatever time unit that fits, and then store it in Excel, in blocks, not in a lengthy line that will be painstakingly long to scroll down to the middle, don't you agree?

If instead of processing with MATLAB you are exporting the raw data to Excel expecting to process data with Excel functions, you may want to consider to process first in MATLAB and the pass to Excel key data, to then plot pie charts, or prepare data for a whatever PowerPoint presentation.

Whatever processing you want to do with Excel functions, it can be done more efficiently in MATLAB.

Bear in mind that you can store directly in .mat

In order to supply something that qualifies to answer, would you consider supplying to this question more detail, like a sample of what kind of processing you want to perform in Excel?

[J Culpepper]

Hey John, I would prefer to save in MATLAB, as I will be doing my post-processing in MATLAB. It's just that whenever I try to save the data in MATLAB, it automatically puts it into Excel. This is the reason I was basing my question around Excel. An example of something I will be doing is taking a waveform, and integrating it in MATLAB then scaling it to match a similar waveform. However, once I can get the data into another program, I hope to be able to do any sort of post-processing typical of laboratory testing.

[J Culpepper]

Would you suggest simply exporting the long string of data into another MATLAB program where I will perform any post-processing? It seems this is a better solution than exporting from MATLAB at all. Thank you for your suggestion.

Answer 1

[Image Analyst]

Did you try this to export into an Excel workbook:

% Create sample data
csvData = sprintf('%.2f', rand());
for k = 1 : 7000
  csvData = sprintf('%s,%.2f', csvData, rand());
end
% Now, csvData is a 35,004 character long character array 
% of numbers separated by commas.
% Now we have sample data and we can begin.

% Now remove commas and transpose.
splitData = strsplit(csvData,',')';
% splitData is a 70001 cell array in a column.
% Write it to Excel into column A
xlswrite('delete me.xlsx', splitData, 'Results', 'A1');
% Now all 7001 numbers will be in rows 1-7001 of the workbook column A.

Answer 2

[Image Analyst]

J, you say "I would prefer to save in MATLAB, as I will be doing my post-processing in MATLAB" so then just use save(), like

% Now remove commas and transpose.
splitData = strsplit(csvData,',')';
% Save to a .mat file in proprietary MATLAB format.
save(someFileName, 'splitData');

[J Culpepper]

This is what I ended up doing, thank you very much for your help, I.A.