How to use the zat.log_to_sparkdf.LogToSparkDF function in zat

def test():
    """Test for LogToSparkDF Class"""
    import os
    from zat.utils import file_utils
    from pyspark.sql import SparkSession

    # Spin up a local Spark Session (with 4 executors)
    spark = SparkSession.builder.master('local[4]').appName('my_awesome').getOrCreate()

    # Grab a test file
    data_path = file_utils.relative_dir(__file__, '../data')
    log_path = os.path.join(data_path, 'ftp.log')

    # Convert it to a Spark DataFrame
    log_to_spark = LogToSparkDF(spark)
    spark_df = log_to_spark.create_dataframe(log_path)

    # Print out the head
    print(spark_df.show())

    # Print out the datatypes
    print(spark_df.printSchema())

    num_rows = spark_df.count()
    print("Number of Spark DataFrame rows: {:d}".format(num_rows))
    columns = spark_df.columns
    print("Columns: {:s}".format(','.join(columns)))

    # Test a bunch
    tests = ['app_stats.log', 'dns.log', 'http.log', 'notice.log', 'tor_ssl.log',
             'conn.log', 'dhcp.log', 'dhcp_002.log', 'files.log',  'smtp.log', 'weird.log',

def log_to_parquet(log_in, parquet_out):
    # Spin up a local Spark Session (with 4 executors)
    spark = SparkSession.builder.master('local[4]').appName('my_awesome').getOrCreate()

    # Use the ZAT class to load our log file into a Spark dataframe (2 lines of code!)
    spark_it = log_to_sparkdf.LogToSparkDF(spark)
    spark_df = spark_it.create_dataframe(log_in)

    # Write it out as a parquet file
    spark_df.write.parquet(parquet_out)
    print('{:s} --> {:s}'.format(log_in, parquet_out))