Export Cell-Data to .xlsx [Saves less data]

Hello Everyone,

I am using following code to export Cell-data to .xlsx file.

filename='(AutoCheck7000).xlsx';
output = [header; Dataset];
xlswrite(filename,output); % Write to the new excel file. 

Where as;

header is cell with 1*13 size

dataset is cell with 189*13 size

I have concentrated them using a variable output but xlswrite is only writing 2 header names instead of all names.

For instance;

the result is:

Whereas, the output-variable is as follows, so .xlsx should show all header names as well:

Any help would be appreciated :-)

Best regard,s

Waqar Ali Memon

[Stephen23]

@Waqar Ali Memon: please upload the variables header and Dataset by clicking the paperclip button.

[Waqar Ali Memon]

Hello Stephen Cobeldick,

I have clipped both Dataset and header variables.

Thank you :-)

[Waqar Ali Memon]

Hello Stephen Cobeldick,

While using following code:

file=strcat(strcat('FileName.xlsx'));
fid = fopen(file, 'wt') ; %%% opens above created CSV file.
fprintf(fid, '%s,', header{1,1:end-1}) ; %%% saves header in it.
fprintf(fid, '%s\n', header{1,end}) ;    %%% moves cursor to next row.
fclose(fid); %%% closes above created CSV file.
xlswrite(filename,Dataset);

I am getting this error:

Error using fprintf

Function is not defined for 'cell' inputs.

Answer 1

[Peter Jarosi]

I recommend converting your cell-data to table and using function writetable().

filename='myExcelFile.xlsx';
myTable = cell2table(Dataset, 'VariableNames', header);
writetable(myTable, filename, 'WriteVariableNames', true);

There is a new function writecell() introduced in R2019a, which can write cell objects into Excel files directly. You may also want to try it. https://www.mathworks.com/help/matlab/ref/writecell.html

[Waqar Ali Memon]

Great, this is also working and it sounds clean as well. But there is only one option :-( to accept the answer. So, I am accepting this and wish you all the best for your Star Badge. You deserve it :-)

Thank you :-)

[Peter Jarosi]

Than you very much!

I appreciate that you accepted this answer.

[Waqar Ali Memon]

Just an update:

Before using above code we need to follow 3 steps suggested by Peter Jarosi in other post.

First Step:

header = cellfun(@char, header, 'un', 0);

Second Step: Construct valid MATLAB identifiers from input strings.

header = matlab.lang.makeValidName(header)

Third Step:

Dataset = cellfun(@num2str, Dataset, 'un', 0);

Finally:

filename='myExcelFile.xlsx';
myTable = cell2table(Dataset, 'VariableNames', header);
writetable(myTable, filename, 'WriteVariableNames', true);

Credit goes to Peter Jarosi :-)

[Peter Jarosi]

Thanks!

I like your more elegant method fixing header problems:

header = matlab.lang.makeValidName(header)

[Waqar Ali Memon]

Hello Peter Jarosi,

Hope you would be doing great:-). I am incurring a little problem here. By exporting a data to table and than writing a table removes pre 0. I mean it saves 0409 to 409. Is there any way to preserve initial zeros?

Second, can we sort the data in columns as well? For instance, it sorts Text data to flush left and numerical data to flush right. Can we initialize a directon for that?

Any help would be appreciated.

Thank you :-)

Answer 2

[Peter Jarosi]

Actually, your Dataset and header need some massage before converting to table:

filename='(AutoCheck7000).xlsx';
load('Dataset.mat');
Dataset = cellfun(@num2str, Dataset, 'un', 0);
load('header.mat');
header = cellfun(@char, header, 'un', 0);
header = strrep(header, '02', 'file_02');
header = strrep(header, ' ', '_');
header = strrep(header, '.', '_');
header = strrep(header, '-', '_');
header = strrep(header, '(', '_');
header = strrep(header, ')', '_');
header = strrep(header, '[', '_');
header = strrep(header, ']', '_');
myTable = cell2table(Dataset, 'VariableNames', header);
writetable(myTable, filename, 'WriteVariableNames', true);

It'll work for you. I've tested.

[Waqar Ali Memon]

Hello Peter Jarosi,

Thank you for the solution, could you please elaborate, why and how to add these messages.

For example, if new header is added, show i need to add extra message ?

Once again, appreciated :-)

[Peter Jarosi]

You're very welcome Waqar Ali Memon!

Massaging header was needed, because it violated the rules of variable names. Quote "A valid variable name begins with a letter and contains not more than namelengthmax characters. Valid variable names can include letters, digits, and underscores." End quote. https://www.mathworks.com/help/matlab/ref/isvarname.html

Your header was in cells of cells format instead of cells of strings format, that's why I applied

header = cellfun(@char, header, 'un', 0);

Your dataset contained mixed numeric and string cells, so I converted all of them to strings.

Dataset = cellfun(@num2str, Dataset, 'un', 0);

Usually it's not necessary but it seemed to be the easiest way to fix your problem.

You may also accept my other answer, because that is more general, so that it might help others to solve their problems (and helps me to achieve rising star badge :-))