In a CSV file I'm loading into Neo4j, a float type column has numbers in some rows, and "NA" in some other rows. This of course results in an error when importing: "Not a number".
I would still like to import this, but with the float stored as a NaN, or omit the property entirely. How can I do this?