the problem with large amount of data is, you don't have an overview of the data.
for example, if you have NaN in your data, and you plot it, it is posible that you don't see where this NaNs are. a good way to find them is using max and min.
if you have some NaNs in your data, and you whant to convert them to zeros (or whatever), use the following command:
data(isnan(data)) = 0; % convert NaN into 0!!!!have a while!