Transpile a SQL Script into PySpark DataFrame API equivalent code

SELECT b.se10,
       b.se3,
       b.se_aggrtr_indctr,
       b.key_swipe_ind
FROM
  (SELECT se10,
          se3,
          se_aggrtr_indctr,
          ROW_NUMBER() OVER (PARTITION BY SE10
                             ORDER BY se_aggrtr_indctr DESC) AS rn,
          key_swipe_ind
   FROM fraud_details_data_whole
   GROUP BY se10,
            se3,
            se_aggrtr_indctr ,
            key_swipe_ind) b
WHERE b.rn<2

from pyspark.sql import SparkSession
from pyspark.sql.window import Window
from pyspark.sql.functions import row_number

# Create a SparkSession
spark = SparkSession.builder \
    .appName("TranspileSQLExample") \
    .getOrCreate()

# Sample data (replace with your actual DataFrame)
data = [
    ("se10_value1", "se3_value1", "aggrtr_value1", "swipe_value1"),
    ("se10_value1", "se3_value2", "aggrtr_value2", "swipe_value2"),
    ("se10_value2", "se3_value3", "aggrtr_value3", "swipe_value3"),
    ("se10_value2", "se3_value4", "aggrtr_value4", "swipe_value4"),
    ("se10_value3", "se3_value5", "aggrtr_value5", "swipe_value5"),
    ("se10_value3", "se3_value6", "aggrtr_value6", "swipe_value6")
]

# Create DataFrame
fraud_details_data_whole = spark.createDataFrame(data, ["se10", "se3", "se_aggrtr_indctr", "key_swipe_ind"])

# Define Window specification for row_number() function
windowSpec = Window.partitionBy("se10").orderBy(fraud_details_data_whole["se_aggrtr_indctr"].desc())

# Add row number column
fraud_details_data_whole = fraud_details_data_whole.withColumn("rn", row_number().over(windowSpec))

# Filter rows where rn < 2
result_df = fraud_details_data_whole.filter("rn < 2")

# Select required columns
result_df = result_df.select("se10", "se3", "se_aggrtr_indctr", "key_swipe_ind")

# Show the result DataFrame
result_df.show()