pyspark dataframe problem

Find pairwise correlations of the numerical columns.
Checking duplicates and filling missing observations with means
Using SparkSQL to find number of distinct restaurants in each zipcode.
You need to run it on PySpark and show results.