Error with Sortrow function in Matlab R2016b

I have 310 data files which are in .cmn format. Each file consists of few paramters, on which the data is collected. When I execute the code in Matlab R2016b, I got the follwoing error, which is as follows.

Error using sortrows (line 62)

COL must be a vector of column indices into X.

Error in sample (line 22)

Cs = sortrows(x,2);

Note: Orginally, the code was written and executed in Matlab R2015, where the code has run successfully and the outputs were shown. Right now, because of unavailiablity of Matlab R2015 version, I am executing the code in Matlab R2016b version. Will this be one of the possible reasons for getting the error? Does Matlab version effect the exection of code? If not, please suggest me a way to clear the error. As per my knowledge, there might be problem with the sortrow function. Please let me know if I am wrong.

• Note: Here is the code attached below. Kindly please make a look over it.
• Note Imp:I am also attaching the screenshot of one file being used, as a sample, to let u know the format of data.

clc; 
clear all; 
close all; 
FileList = dir('*.Cmn'); 
N = size(FileList,1); 
for k = 1:N    %get the file name:    
filename1{k}= FileList(k).name;    
disp(filename1); 
end  
 % In sprintf(‘name of the first file of the year in the folder’) 
 
for k = 1:N,     
filename=filename1{k};     
myfilename = sprintf('SCND004-2016-01-04%d.Cmn', k);      
mydata{k} = importdata(filename); 
end 
 

%TEC processing 
 
AA=[] 
for i=1:N, 
x=mydata{:,i}; 
Cs = sortrows(x,2); 
idx=find(Cs(:,2)>0); 
CC=Cs(idx,:); 
C1=CC(:,9); 
en=C1(1:end/24);
en1=C1(1:end); 
cc=length(en1); 
cc1=13750-cc;  
b = cat(1, C1, zeros(cc1, 1));  
c = reshape(b,550,25);  
C=mean(c) ;  
C=C(:,1:24); 
AA=[AA C]; 
End 
 
AIT2016=reshape(AA,24,N); 
singleTec=AIT2016(:,2) 
 
 % In place of 2 event day can be replaced so can observe for desired day 
 
plot (AIT2016); 
save ('AIT2016','AIT2016'); % AIT2017 for 2017 Data 

[Walter Roberson]

At some point one of the files is read as only one column.

[mounica penmetsha]

I would like to thank you for the quick reply @Walter Roberson. If you don't mind, could you please help me a way, to solve the above error.

[Guillaume]

Note that the problem has nothing to do with your version of matlab. It's a bug in your code that you're only finding now because one of the input file is somehow different from the previous files you've used.

Answer 1

[Walter Roberson]

Code attached, rewritten to use tables and simplified. Requires R2013b or later.

[mounica penmetsha]

@Walter Roberson, thank you for taking the time and effort for rewriting the code. I just executed the code in R2016b and I got the following error.

Error using readtable (line 197)

The file extension '.Cmn' is not recognized.

Error in walterrobersoncode (line 9)

mydata{k} = readtable(filename);

[Guillaume]

Add 'FileType', 'text' to that readtable call, so:

mydata{k} = readtable(filename, 'FileType', 'text');

[Walter Roberson]

Guillaume's solution is a good one.

[mounica penmetsha]

Thank you @Guillaume. The error has been resolved.

Answer 2

[Jan]

Start with omitting the brute clearing header:

clc; 
clear all; 
close all; 

It is a waste of time to kill everything. Better use functions to keep the workspace clean. All you need is to start the code with the keyword "function".

Now add this line:

x=mydata{:,i};  % from your code
if size(x, 2) < 2
    fprintf(2, '%s contains unexpected data\n', filename1{k});
    continue;
end

By the way, this can be simplified:

N = size(FileList,1); 
for k = 1:N    %get the file name:    
filename1{k}= FileList(k).name;    
disp(filename1); 
end  

to

filename1 = {FileList.name};
fprintf('%s\n', filename1{:});