I am trying to run a Python script using Spark that read from a file and classify the words into positive and negative state, then write them to two files (tf_pos.csv and tf