You are working with a large-scale data processing project that uses Hadoop for storage and PySpark for data processing. Explain how you would set up a PySpark job to read data from HDFS, perform a transformation to filter out records with missing values, and then write the cleaned data back to HDFS. Provide a sample PySpark code snippet to demonstrate this process.