How can we write data into csv file with column labels included ?

I want to write the data into csv file with column labels. I have illustrated the code below.

chead = {'a','b','c','d','e','f','g','h','i','k','l'};
data1 = ones(1001,4);
data2 = ones(125,7);
data = {data1,data2};
R = [chead;data];
writecell(R,'file.csv');

But this doesn't work the way i expected. This was writing the data like:

a b c .......

1 1 1 1 .... 1 1 1 1 .... 1 1 1 1....

instead of

a b c .......

1 1 1

1 1 1

1 1 1

1 1 1

. . .

. . .

If anyone knows please tell the answer.

[KSSV]

REad about writetable.

[Mahesh Babu Dhanekula]

It doesn't support vectors having variable row length. Can you tell me any workaround method for using this command ?

Answer 1

[dpb]

...
data1 = ones(1001,4);
data2 = ones(125,7);
data = {data1,data2};
...

You have created a disparate-sized array of 125 rows of seven variables and the remainder with only four -- this can't be put into a rectangular array to write a regular .csv file. writecell did what you asked it to do, output the content of the cell array you built; you just didn't build a representation of what you (apparently) intended.

What do you intend the content of the file to be for the shorter records -- missing values or shorter records?

I've not tested if writecell will build a file -- well, let's just see:

>> writecell(cell(2,4))
>> type cell.txt

,,,
,,,
>> 

Ah! Indeed it will, if you convert your numeric array by num2cell to a cell array of one element to cell, and add the empty cells where needed, joy wil ensue.

ADDENDUM:

While it's relatively trivial to write using fprintf, you can try something like

names={'a','b','c'};    % build a set of variable names
writecell(names)        % write to names.txt (pick your name as desired)
c= [ones(2,3) zeros(2,4)];          % longer rows
save names.txt c -ascii -append     % add to existing file
c= ones(2,3);                       % short rows
save names.txt c -ascii -append     % ditto

results in

>> type names.txt

a,b,c
   1.0000000e+00   1.0000000e+00   1.0000000e+00   0.0000000e+00   0.0000000e+00   0.0000000e+00   0.0000000e+00
   1.0000000e+00   1.0000000e+00   1.0000000e+00   0.0000000e+00   0.0000000e+00   0.0000000e+00   0.0000000e+00
   1.0000000e+00   1.0000000e+00   1.0000000e+00
   1.0000000e+00   1.0000000e+00   1.0000000e+00
>> 

[Mahesh Babu Dhanekula]

but i am not interested in having additional zeros in columns. so i just wanted this to work as i expected.

[dpb]

You'll have to write the file specifically with low-level calls, then. There is no prepared function to write an irregular file of that format.

You'll run into trouble trying to read the file back again, anyway, unless an application is aware of this issue going in.

Answer 2

[giancarlo maldonado cardenas]

you can do it with this

A=[4 5 6;7 8 9]

T = array2table(A)

T.Properties.VariableNames(1:3) = {'x_axis','y_axis','z_axis'}

writetable(T,'file1.csv')

[dpb]

That handles only regular arrays; not the variable-sized cells of the original Q?.