Reputation: 6668
I have a large cell array. I have 3 columns which are numerical. Some of the rows in these columns are NaN. This causes an issue when I upload the data to SQL Server (I do not wish to use the matlab function insert).
So far I have tried the below,
export_full = strrep(export_full(:, 6:9),'NaN','NULL');
which return the error message
Cell elements must be character arrays.
Can I use indexing to solve this problem if so how? Performance would be an issue too as its quite a large cell array.
Edit
My cell array is a 10000 x 10 cell column 1 is string column 2 is string column 3 is date column 4 to 10 is numerical
Upvotes: 1
Views: 861
Reputation: 45752
I think this should do the trick (credit goes to: http://www.mathworks.com/matlabcentral/newsreader/view_thread/283735)
ind = cellfun(@(x) any(isnan(x(:))), export_full)
export_full(ind) = {'NULL'};
Upvotes: 1