josep2

  val result = g.stronglyConnectedComponents.maxIter(5).run()
  result.show()

+---+-------+---+--------+----------+---------+
| id|   name|age|    cash|     fruit|component|
+---+-------+---+--------+----------+---------+
|  a|  Alice| 34|     234|    Apples|        0|
|  g|  Gabby| 60|   23433|   Oranges|        6|
|  f|  Fanny| 36|     333|    Apples|        5|
|  b|    Bob| 36|23232323|   Bananas|        1|

josep2

val results = g.pageRank.resetProbability(0.15).tol(0.05).run()
// resetProbability and TOL are convergence parameters
results.edges.select("src", "dst", "weight").show()

+---+---+------+
|src|dst|weight|
+---+---+------+
|  d|  a|   1.0|
|  a|  b|   0.5|
|  a|  e|   0.5|

josep2

// Borrowed some parts from the GraphFrame docs for my blog: https://graphframes.github.io/user-guide.html
import org.graphframes._

val v = sqlContext.createDataFrame(List(
  ("a", "Alice", 34, 234, "Apples"),
  ("b", "Bob", 36, 23232323, "Bananas"),
  ("c", "Charlie", 30, 2123, "Grapefruit"),
  ("d", "David", 29, 2321111, "Bananas"),
  ("e", "Esther", 32, 1, "Watermelon"),
  ("f", "Fanny", 36, 333, "Apples" ),

josep2

// From https://github.com/amplab/succinct
import edu.berkeley.cs.succinct.sql._

// Create a schema
val citySchema = StructType(Seq(
  StructField("Name", StringType, false),
  StructField("Length", IntegerType, true),
  StructField("Area", DoubleType, false),
  StructField("Airport", BooleanType, true)))

josep2

import edu.berkeley.cs.succinct._


val conf = new SparkConf().setAppName("Ranking Example")

val sc = new SparkContext(conf)


// A large file of raw hip hop lyrics ~ 100 GB
val hipHopRDD = sc.textFile("/hiphopcorpus").map(_.getBytes)

josep2

import edu.berkeley.cs.succinct._


val conf = new SparkConf().setAppName("Ranking Example")

val sc = new SparkContext(conf)


// A large file of raw hip hop lyrics ~ 100 GB
val hipHopRDD = sc.textFile("/hiphopcorpus").map(_.getBytes)

josep2

  

  // Create a General linear model parameters object from the Gaussian family in this case

  val glmParams = new GLMParameters(Family.gaussian)
  
  // Give it the data to train. You can get around using the key attribute by Lock and Update
  glmParams._train = h2oData.key
  // Set your dependent variable
  glmParams._response_column = "Some Dependent Variable"

josep2

+-------+-----------+
| x     |    y      |
+-------+-----------+
|9285958|0.492819875|
|6295780|0.334126593|
| 549166|0.029145073|
| 408830|0.021697228|
| 362166|0.019220699|
| 305267|0.016200983|
| 270418|0.014351493|

josep2

import org.apache.spark.h2o.H2OContext
import org.apache.spark.sql.SparkSession


object PartyStarted extends App {

  val sparkSession = SparkSession.builder
    .master("local[*]")
    .appName("Try H2o")
    .config("spark.sql.crossJoin.enabled", "true")

josep2

"ai.h2o" %% "sparkling-water-core" % "2.0.3"