Why does PySpark tells key error on a column name even though the column is present in the dataframe?

Question

I am trying to save a dataframe into snowflake after reading a parquet file from HDFS as below. Before loading it into the snowflake table, an ID column is being added to it