#!/usr/bin/env python3
# coding: utf-8

import json
import pandas
import findspark
findspark.init()
import pyspark
import pyspark.sql
import sys
import gzip
import os

import pyspark.sql.functions as F
from pyspark import AccumulatorParam
from pyspark.sql.functions import lit
from pyspark.sql import Window
from pyspark.sql.types import StructType, LongType, StringType, ByteType

def main():
    global DIR
    global NAME
    global launched

    if len(sys.argv) != 4:
        print(sys.argv[0] + " {cluster} {tmpdir} {maxram}")
        sys.exit()

    cluster=sys.argv[1]
    DIR = os.path.dirname(__file__)
    NAME = "table-iii-" + cluster

    if os.path.exists(DIR + "/" + NAME + "-working") or os.path.exists(DIR + "/" + NAME + ".parquet"):
        print("already launched")
        launched = True
        sys.exit()

    os.system("touch " + DIR + "/" + NAME + "-working")

    spark = pyspark.sql.SparkSession.builder \
      .appName(NAME) \
      .config("spark.driver.maxResultSize", "32g") \
      .config("spark.local.dir", sys.argv[2]) \
      .config("spark.driver.memory", sys.argv[3]) \
      .getOrCreate()
    sc = spark.sparkContext

    df = spark.read.parquet(DIR + "/bigtable-" + cluster + ".parquet")

#     df = df.groupBy("jobid").agg(
#         F.expr("sum(count_0)").alias('count_0'),
#         F.expr("sum(count_1)").alias('count_1'),
#         F.expr("sum(count_2)").alias('count_2'),
#         F.expr("sum(count_3)").alias('count_3'),
#         F.expr("sum(count_4)").alias('count_4'),
#         F.expr("sum(count_5)").alias('count_5'),
#         F.expr("sum(count_6)").alias('count_6'),
#         F.expr("sum(count_7)").alias('count_7'),
#         F.expr("sum(count_8)").alias('count_8'),
#         F.expr("sum(count_9)").alias('count_9'),
#         F.expr("sum(count_10)").alias('count_10'))
# 
#     usage_schema = StructType() \
#         .add("jobid", StringType(), False) \
#         .add("job_term", LongType(), False) \
#         .add("task_count", LongType(), False)    
#     dfj = spark.read.format("csv") \
#         .option("header", False) \
#         .schema(usage_schema) \
#         .load("/home/claudio/google_2019/thesis_queries/figure_9/" + cluster + "_task_count")
#     dfj = dfj.select(["jobid", "job_term"])
# 
#     df = df.join(dfj, "jobid")
    df = df.withColumn("count_sum", 
        df["count_0"] + \
        df["count_1"] + \
        df["count_2"] + \
        df["count_3"] + \
        df["count_4"] + \
        df["count_5"] + \
        df["count_6"] + \
        df["count_7"] + \
        df["count_8"] + \
        df["count_9"] + \
        df["count_10"])

    df = df.groupBy("task_term") \
        .agg(F.expr("percentile(count_sum, array(0.95))")[0].alias('%95'),
        F.expr("avg(count_sum)").alias('mean'),
        F.expr("avg(count_0)").alias('avg_count_0'),
        F.expr("avg(count_1)").alias('avg_count_1'),
        F.expr("avg(count_2)").alias('avg_count_2'),
        F.expr("avg(count_3)").alias('avg_count_3'),
        F.expr("avg(count_4)").alias('avg_count_4'),
        F.expr("avg(count_5)").alias('avg_count_5'),
        F.expr("avg(count_6)").alias('avg_count_6'),
        F.expr("avg(count_7)").alias('avg_count_7'),
        F.expr("avg(count_8)").alias('avg_count_8'),
        F.expr("avg(count_9)").alias('avg_count_9'),
        F.expr("avg(count_10)").alias('avg_count_10'))

    df.repartition(1).write.option("header", True).csv(DIR + "/" + NAME + ".csv")

if __name__ == "__main__":
    launched = False
    try:
        main()
    finally:    
        if not launched:
            os.system("rm -v " + DIR + "/" + NAME + "-working")

# vim: set ts=4 sw=4 et tw=120: