I have a below input:

I have loaded a set of JSON objects in a Spark data frame called t. Inner objects / arrays are parsed and queryable.

I’m looking to melt my pyspark dataframe in a customized manner.
My dataframe looks like below

Hi I am trying to save the content of a column(contains xml information) from a data frame to an xml file. Whenever i run my program i get the error “col should be Column” indicating that when i call my function that saves the content of the column to an xml file, it is saying that it doesn’t recognize the column as a column. I have tried using an UDF with it but get the same error. I upload a picture of my code below, any suggestion would help!