When using a da_datastore for data management, the special convention for floating point not a number (NaN) can be used to denote missing
data. See Data Management Functionalities for further details.
In order to provide the best possible performance, the algorithmic functions will not automatically check for
NaN data. If a NaN is passed into an algorithmic function, its behavior is undefined.
It is therefore the user’s responsibility to ensure data is sanitized before passing it to one of the algorithms (for example, by using
da_data_select_non_missing(), by calling da_check_data_s(), or by setting the check data option in the algorithmic APIs that use handles).