I have a numeric column which contains entries for some but not all rows. The missing elements start out as NULL. But if I delete values in the web UI then they seem to become the empty string instead. This then causes downstream problems when accessing the data from the python API since the column is no longer pure numeric/null. Is this expected? Can anyone suggest a workaround?
So far I am resorting to using the API to
- find rows containing the empty string via SQL query
- update those rows with numeric NA values
Thank you! Greg.