I have a large database of customer demands with additional information (address, name etc) that is preprocessed in FME before it is imported into a model using modelbuilder. Some customers are missing some data (address/name etc.), how should I preprocess this data to import a valid database into the model using modelbuilder? I have the option to treat the missing data in 3 ways:
- Empty (attribute exists as an empty string)
- Null (attribute exists with value null)
- Missing (attribute does not exist)
Currently the missing attributes are stored as null, is modelbuilder able to handle null values?
Hello Joeri,
Rather than keeping the values as "null" try and have a common attribute for such data like a text string "NA" or a numeric value "0" which you help you map this field when importing via ModelBuilder. This would ensure that some value is imported when running the ModelBuilder and you would be able to easily identify it.
In what format are you building this database? Shapefiles? Spreadsheets?
Regards,
Yashodhan Joshi
Answer Verified By: Joeri Legierse
NA for text and 0 for integers & doubles would work, might give this a try. Format is .accdb because shapefile is limited in file size (2gb) and because I first have to import the database with the tap xy coordinates, run loadbuilder to create taps & laterals and then update the xy-coordinates to the meter coordinates.