Creating PySpark UDFs from python method with numpy array input, to calculate and return a single float value

Question

As input I have a csv file with float values in it.

spark_df = spark.read.option("header", "false").csv("../float_values.csv")

df =