cellfun(@isnan) and erasing NaN

Dear all,

I have a matrix like this in excel file:

A=[a, 1, NaN,  1,  1, 1, 1;...
    b,  2,  NaN,  NaN,  NaN,  NaN,  NaN;...
    c,  3,  NaN,  3,  3,  3,  3];

and I want to erase the rows in which has NaN starting from fourth column. In another words I would like to have

B=[a, 1, NaN,  1,  1, 1, 1;...
   c,  3,  NaN,  3,  3,  3,  3];

I used this:

[ndata, text, a] = xlsread('test.xls');
xlswrite('newfile.xls', a(~any(cellfun(@isnan,a), 2), :));

But it works just in cases of having numbers in the excel file. But with the above example ,in which I have a, b,and c, I receive this error:

Error using cellfun Non-scalar in Uniform output, at index 1, output 1. Set 'UniformOutput' to false.

Or if you have other idea to do this operation please let me know.

Answer 1

[Sean de Wolski]

Close!

A={'a', 1, NaN,  1,  1, 1, 1;...
 'b',  2,  NaN,  NaN,  NaN,  NaN,  NaN;...
 'c',  3,  NaN,  3,  3,  3,  3};

B = A(~any(cellfun(@isnan,A(:,4:end)),2),:)

[Victor]

can you suggest writing that in form of "xlswrite"!!

[ndata, text, a] = xlsread('test.xls');

xlswrite('newfile.xls', a(~any(cellfun(@isnan,a(:,4:end)),2),:));

Because it gives:

Error using cellfun Non-scalar in Uniform output, at index 2, output 1. Set 'UniformOutput' to false.

[Sean de Wolski]

Break it into pieces.

What does whos() return after running the first line?

Then run the line I have. Does it work?

If not, why not? If so, then worry about xlswrite()

[Victor]

Thanks, now it works.

Answer 2

[Azzi Abdelmalek]

Use

cellfun(@isnan,a,'UniformOutput',false)

Which means

xlswrite('newfile.xls', a(~any(cellfun(@isnan,a,'UniformOutput',false), 2), :));

[Victor]

Thanks, but:

Undefined function 'any' for input arguments of type 'cell'.

[Azzi Abdelmalek]

Then use

xlswrite('newfile.xls', a(~any(cell2mat(cellfun(@isnan,a,'un',0)), 2), :));

[Victor]

Thanks

xlswrite('newfile.xls', a(~any(cell2mat(cellfun(@isnan,a,'un',0)), 2), :));

Error using cat Dimensions of matrices being concatenated are not consistent.

[Azzi Abdelmalek]

I've tried to fellow your code, now try this

A={'a', 1, NaN,  1,  1, 1, 1
    'b',2,NaN,NaN,NaN,NaN,NaN
    'c',3,NaN,3,3,3,3}
b=A(:,4:end)
A=A(~all(cellfun(@isnan,b),2),:)

[Victor]

I think I am doing something wrong... When I read the file with "xlsread" that is in not in the format of cell. By the way I tried to use "cell2mat" so that I can write it with xlswrite, but " All contents of the input cell array must be of the same data type."

a=cell2mat(a)

xlswrite('newfile.xls',a);

[Azzi Abdelmalek]

You have not to use cell2mat before xlswrite

[Victor]

Thanks Azzi, but to be honest, I do not think with this method, I can read the file, do the manipulation and then write it back.