I have been trying to implement a udf in pyspark out of a py function as follows:
It takes a bin model I previously trained.
def sentiment(frase): clas
