[Num,Txt,R​aw]=xlsrea​d......

I used [Num,Txt,Raw]=xlsread to import a matrix containing both numerical and text data and Raw is a 2000*8 cell. When I try to compute eg mean(Raw(.,7)) the following message appears:

??? Undefined function or method 'sum' for input arguments of type 'cell'.

Error in ==> mean at 29 y = sum(x)/size(x,dim);

How can I convert cel into num values?

Thanks

[C.J. Harris]

Why don't you use the numeric data to compute the mean etc?

[Cristina]

Because I have to make computations conditional to a list of variables.... e.g. Raw contains a colums country, rating... and I have to compute let's say the mean only if country=IT and rating=BBB. So I need both numerical and text data...

Answer 1

[Kye Taylor]

To convert cell to matrix use the cell2mat functions. For example,

mean(cell2mat(Raw(:,7)))

Answer 2

[Andrei Bobrov]

sum for arguments of type double

eg

mean(Num(:,7));

[Cristina]

ok but this is not a conditional mean...I need to compute the mean if eg country='IT'

Answer 3

[Peter Perkins]

Cristina, if you have access to the Statistics Toolbox, you might find that using a dataset array is helpful the kind of thing you seem to be describing. You can read a spreadsheet into a dataset array, and then, for example, compute the mean of one variable conditional on values of other numeric or non-numeric variables.

This example on the File Exchange

http://www.mathworks.com/matlabcentral/fileexchange/24267-an-introduction-to-dataset-arrays

seems more or less the kind of thing you are describing.

Answer 4

[Syed Zubair shah]

what does it means of Num , txt , Raw ... can we just assign it by T =xlsread(filename) ... if yes then how we call specific column then like F is a column of 1600 rows ?

[Num,Txt,R​aw] = xlsread(filename)

[Image Analyst]

[Num,Txt,Raw] = xlsread(filename, sheetName, 'F1:F1600');