1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
|
package com.yahoo.example.blog
import org.apache.spark.sql.{SparkSession, DataFrame}
import org.apache.spark.sql.functions.udf
import org.apache.spark.sql.functions._
class SplitFullSetIntoTrainAndTestSets(val ss: SparkSession) {
private def loadAndSimplifyFullDataset(input_file_path: String): DataFrame = {
// Load full dataset
val full_dataset = ss.read.json(input_file_path)
val full_dataset_simple = full_dataset.select(col("post_id"), size(col("likes")).as("number_likes"), col("likes"))
full_dataset_simple
}
private def splitSimplifiedDatasetIntoTrainAndTestSets(full_dataset_simple: DataFrame,
test_perc_stage1: Double,
test_perc_stage2: Double,
seed: Int): Array[DataFrame] = {
// Set some blog posts aside to be present only on the test set
var sets = full_dataset_simple.randomSplit(Array(1 - test_perc_stage1, test_perc_stage1), seed)
val training_set = sets(0)
val training_set_null = training_set.filter("number_likes = 0")
var training_set_exploded = training_set.select(col("post_id"), explode(col("likes")).as("likes_flat"))
training_set_exploded = training_set_exploded.select("post_id", "likes_flat.uid")
val test_set = sets(1)
val test_set_null = test_set.filter("number_likes = 0")
var test_set_exploded = test_set.select(col("post_id"), explode(col("likes")).as("likes_flat"))
test_set_exploded = test_set_exploded.select("post_id", "likes_flat.uid")
// randomly move some (post_id, uid) from training set to test set
sets = training_set_exploded.randomSplit(Array(1 - test_perc_stage2, test_perc_stage2), seed)
training_set_exploded = sets(0)
val additional_test_set_exploded = sets(1)
test_set_exploded = test_set_exploded.union(additional_test_set_exploded)
// concatenate exploded set with null set
val getNull = udf(() => None: Option[String])
training_set_exploded = training_set_exploded.union(training_set_null.select("post_id").withColumn("uid", getNull()))
test_set_exploded = test_set_exploded.union(test_set_null.select("post_id").withColumn("uid", getNull()))
Array(training_set_exploded, test_set_exploded)
}
def run(input_file_path: String, test_perc_stage1: Double, test_perc_stage2:Double, seed: Int): Array[DataFrame] = {
val full_dataset_simple = loadAndSimplifyFullDataset(input_file_path)
splitSimplifiedDatasetIntoTrainAndTestSets(full_dataset_simple,
test_perc_stage1,
test_perc_stage2,
seed)
}
}
object SplitFullSetIntoTrainAndTestSets {
def writeTrainAndTestSetsIndices(train_and_test_sets: Array[DataFrame], output_path: String): Unit = {
val training_set_exploded = train_and_test_sets(0)
val test_set_exploded = train_and_test_sets(1)
// Write to disk
training_set_exploded.rdd.map(x => x(0) + "\t" + x(1)).saveAsTextFile(output_path + "/training_set_ids")
test_set_exploded.rdd.map(x => x(0) + "\t" + x(1)).saveAsTextFile(output_path + "/testing_set_ids")
}
}
|
