Compulsory input and output validation

[sebastian-c]

Input

1. Your input data should have NAs scrubbed from it
2. Your input should have a decisive way of handling M flags

Output

1. Your output should have no duplicate values
2. Your output should have no missing values unless you mean to erase those values
3. Your output should have no keys that are not present in the destination dataset

[mkao006]

Rather then give outlines, these should be functionalities in the faoswsProcessing and faoswsEnsure package.

e.g. The preProcessing function in the faoswsProcessing package converts 0 to NA when the observation status flag is M, the ensureCorrectMissingValue checks whether all values which has observation status flag M are either 0 or NA.