NaN values in timetable - how to calculate Nash–Sutcliffe model efficiency coefficient?

I would like the calculate the Nash–Sutcliffe model efficiency coefficient between various variables in a dataset. However a number of rows do contain missing data so I would like to exclue them from the calculate too. Is there a function that will allow this in matlab?

[Jan]

So the actual problem is how to remove rows with missing data? Then please post an example, which shows, how "missing" data are represented.

Answer 1

[Star Strider]

If the ‘missing’ data are NaN entries, this works:

data = rand(10,3);                                  % Create Array
data(randi(30,1,5)) = NaN;                          % Create Missing Data
rowidx = ~any(isnan(data),2);                       % Rows Without ‘NaN’ Entries
data_new = data(rowidx,:);                          % Matrix With No ‘NaN’ Values

If the missing entries are in a cell array with empty cells, this works:

data = num2cell(rand(10,3));                        % Create Cell Array
data(randi(30,1,5)) = {[]};                         % Create Missing Data
rowidx = ~any(cellfun(@isempty,data),2);            % Rows Without Empty Cells
data_new = data(rowidx,:);                          % Matrix With No Empty Cells

.