Problems when I pass Null values to Python RRS feed

  • Question

  • Hello,

    I'm using ADLA with Python scripts, and regarding the documentation:

    U-SQL Nulls are converted to and from Pandas NA values

    But if I pass null values in an int column to python, this null values are updated to 0, so this imply that I can not distinguish between a null value (to do  for example an interpolation) value with a real zero.

    Any help?

    Friday, September 14, 2018 2:27 PM

All replies

  • Hi there,

    It's a workaround, could you pass in as string and convert later, use length to determine nullability?

    In c# it would be "int?" data type.



    Andrew Sears

    Friday, September 14, 2018 8:15 PM
  • Hi, Andrew

    Thanks a lot, yes, I know, i tried that possibility and it worked, but i would like to know whether I'm doing something wrong or is there a documentation error, as it seems.


    Saturday, September 15, 2018 9:36 AM