Code a simple PySpark job to read a JSON file, filter records, and write output in Parquet format.

Production-grade example (with schema, error handling):

from pyspark.sql import SparkSession
from pyspark.sql.functions import col
spark = SparkSession.builder.appName("json_to_parquet").getOrCreate()
# Provide schema to avoid inference cost on large reads
df = spark.read.schema("id INT, status STRING, amount DOUBLE") \
    .json("s3://bucket/input/*.json")
filtered = df.filter((col("status") == "active") & (col("amount") > 0))
filtered.write.mode("overwrite") \
    .parquet("s3://bucket/output/")

Why schema: Inference scans data; explicit schema avoids that on large files. Scalability trade-offs: *.json = many small files = many partitions; coalesce before write to avoid small-file problem. Cost implications: Filter early reduces bytes processed; partition output by date if downstream queries filter by date.