Scala to Python - pairRdd folder

pairRdd/aggregation/combinebykey/AverageHousePriceSolution.py

@@ -0,0 +1,21 @@
+from pyspark import SparkContext
+
+if __name__ == "__main__":
+
+    sc = SparkContext("local", "AverageHousePrice")
+    sc.setLogLevel("ERROR")
+
+    lines = sc.textFile("in/RealEstate.csv")
+    cleanedLines = lines.filter(lambda line: "Bedrooms" not in line)
+
+    housePricePairRdd = cleanedLines.map(lambda line: (line.split(",")[3], float(line.split(",")[2])))
+
+    createCombiner = lambda x: (1, x)
+    mergeValue = lambda avgCount, x: (avgCount[0] + 1, avgCount[1] + x)
+    mergeCombiners = lambda avgCountA, avgCountB: (avgCountA[0] + avgCountB[0], avgCountA[1] + avgCountB[1])
+
+    housePriceTotal = housePricePairRdd.combineByKey(createCombiner, mergeValue, mergeCombiners)
+
+    housePriceAvg = housePriceTotal.mapValues(lambda avgCount: avgCount[1] / avgCount[0])
+    for bedrooms, avgPrice in housePriceAvg.collect():
+        print("{} : {}".format(bedrooms, avgPrice))

pairRdd/aggregation/reducebykey/WordCount.py

@@ -0,0 +1,14 @@
+from pyspark import SparkContext
+
+if __name__ == "__main__":
+
+    sc = SparkContext("local", "wordCounts")
+    sc.setLogLevel("ERROR")
+
+    lines = sc.textFile("in/word_count.text")
+    wordRdd = lines.flatMap(lambda line: line.split(" "))
+    wordPairRdd = wordRdd.map(lambda word: (word, 1))
+
+    wordCounts = wordPairRdd.reduceByKey(lambda x, y: x + y)
+    for word, count in wordCounts.collect():
+        print("{} : {}".format(word, count))

pairRdd/aggregation/reducebykey/housePrice/AverageHousePriceProblem.scala → pairRdd/aggregation/reducebykey/housePrice/AverageHousePriceProblem.py

@@ -1,11 +1,10 @@
-package com.sparkTutorial.pairRdd.aggregation.reducebykey.housePrice
+from pyspark import SparkContext
 
-object AverageHousePriceProblem {
+if __name__ == "__main__":
 
-  def main(args: Array[String]) {
-
-    /* Create a Spark program to read the house data from in/RealEstate.csv,
-       output the average price for houses with different number of bedrooms.
+    '''
+    Create a Spark program to read the house data from in/RealEstate.csv,
+    output the average price for houses with different number of bedrooms.
 
     The houses dataset contains a collection of recent real estate listings in San Luis Obispo county and
     around it. 
@@ -31,8 +30,6 @@ def main(args: Array[String]) {
        (2, 325000)
        ...
 
-       3, 1 and 2 mean the number of bedrooms. 325000 means the average price of houses with 3 bedrooms is 325000.
-     */
-  }
+    3, 1 and 2 mean the number of bedrooms. 325000 means the average price of houses with 3 bedrooms is 325000.
 
-}
+    '''

pairRdd/aggregation/reducebykey/housePrice/AverageHousePriceSolution.py

@@ -0,0 +1,24 @@
+from pyspark import SparkContext
+
+if __name__ == "__main__":
+
+    sc = SparkContext("local", "avgHousePrice")
+    sc.setLogLevel("ERROR")
+
+    lines = sc.textFile("in/RealEstate.csv")
+    cleanedLines = lines.filter(lambda line: "Bedrooms" not in line)
+
+    housePricePairRdd = cleanedLines.map(lambda line: \
+        (line.split(",")[3], (1, float(line.split(",")[2]))))
+
+    housePriceTotal = housePricePairRdd \
+        .reduceByKey(lambda x, y: (x[0] + y[0], x[1] + y[1]))
+
+    print("housePriceTotal: ")
+    for bedroom, total in housePriceTotal.collect():
+        print("{} : {}".format(bedroom, total))
+
+    housePriceAvg = housePriceTotal.mapValues(lambda avgCount: avgCount[1] / avgCount[0])
+    print("\nhousePriceAvg: ")
+    for bedroom, avg in housePriceAvg.collect():
+        print("{} : {}".format(bedroom, avg))

pairRdd/aggregation/reducebykey/housePrice/AvgCount.py

@@ -0,0 +1,7 @@
+class AvgCount():
+
+    def __init__(self, count: int, total: float):
+        self.count = count
+        self.total = total
+
+

pairRdd/create/PairRddFromRegularRdd.py

@@ -0,0 +1,12 @@
+from pyspark import SparkContext
+
+if __name__ == "__main__":
+
+    sc = SparkContext("local", "create")
+    sc.setLogLevel("ERROR")
+
+    inputStrings = ["Lily 23", "Jack 29", "Mary 29", "James 8"]
+    regularRDDs = sc.parallelize(inputStrings)
+
+    pairRDD = regularRDDs.map(lambda s: (s.split(" ")[0], s.split(" ")[1]))
+    pairRDD.coalesce(1).saveAsTextFile("out/pair_rdd_from_regular_rdd")

pairRdd/create/PairRddFromTupleList.py

@@ -0,0 +1,11 @@
+from pyspark import SparkContext
+
+if __name__ == "__main__":
+
+    sc = SparkContext("local", "create")
+    sc.setLogLevel("ERROR")
+
+    tuples = [("Lily", 23), ("Jack", 29), ("Mary", 29), ("James", 8)]
+    pairRDD = sc.parallelize(tuples)
+
+    pairRDD.coalesce(1).saveAsTextFile("out/pair_rdd_from_tuple_list")

pairRdd/filter/AirportsNotInUsaProblem.py

@@ -0,0 +1,20 @@
+from pyspark import SparkContext
+
+if __name__ == "__main__":
+
+    '''
+    Create a Spark program to read the airport data from in/airports.text;
+    generate a pair RDD with airport name being the key and country name being the value.
+    Then remove all the airports which are located in United States and output the pair RDD to out/airports_not_in_usa_pair_rdd.text
+
+    Each row of the input file contains the following columns:
+    Airport ID, Name of airport, Main city served by airport, Country where airport is located,
+    IATA/FAA code, ICAO Code, Latitude, Longitude, Altitude, Timezone, DST, Timezone in Olson format
+
+    Sample output:
+
+    ("Kamloops", "Canada")
+    ("Wewak Intl", "Papua New Guinea")
+    ...
+
+    '''

pairRdd/filter/AirportsNotInUsaSolution.py

@@ -0,0 +1,16 @@
+from pyspark import SparkContext
+from commons.Utils import Utils
+
+if __name__ == "__main__":
+
+    sc = SparkContext("local", "airports")
+    sc.setLogLevel("ERROR")
+
+    airportsRDD = sc.textFile("in/airports.text")
+
+    airportPairRDD = airportsRDD.map(lambda line: \
+        (Utils.COMMA_DELIMITER.split(line)[1],
+         Utils.COMMA_DELIMITER.split(line)[3]))
+    airportsNotInUSA = airportPairRDD.filter(lambda keyValue: keyValue[1] != "\"United States\"")
+
+    airportsNotInUSA.saveAsTextFile("out/airports_not_in_usa_pair_rdd.text")

pairRdd/groupbykey/AirportsByCountryProblem.py

@@ -0,0 +1,23 @@
+from pyspark import SparkContext
+
+if __name__ == "__main__":
+
+    '''
+    Create a Spark program to read the airport data from in/airports.text,
+    output the the list of the names of the airports located in each country.
+
+    Each row of the input file contains the following columns:
+    Airport ID, Name of airport, Main city served by airport, Country where airport is located, IATA/FAA code,
+    ICAO Code, Latitude, Longitude, Altitude, Timezone, DST, Timezone in Olson format
+
+    Sample output:
+
+    "Canada", ["Bagotville", "Montreal", "Coronation", ...]
+    "Norway" : ["Vigra", "Andenes", "Alta", "Bomoen", "Bronnoy",..]
+    "Papua New Guinea",  ["Goroka", "Madang", ...]
+    ...
+
+    '''
+
+
+

pairRdd/groupbykey/AirportsByCountrySolution.py

@@ -0,0 +1,18 @@
+from pyspark import SparkContext
+from commons.Utils import Utils
+
+if __name__ == "__main__":
+
+    sc = SparkContext("local", "airports")
+    sc.setLogLevel("ERROR")
+
+    lines = sc.textFile("in/airports.text")
+
+    countryAndAirportNameAndPair = lines.map(lambda airport:\
+         (Utils.COMMA_DELIMITER.split(airport)[3],
+          Utils.COMMA_DELIMITER.split(airport)[1]))
+
+    airportsByCountry = countryAndAirportNameAndPair.groupByKey()
+
+    for country, airportName in airportsByCountry.collectAsMap().items():
+        print("{}: {}".format(country,list(airportName)))