ruby-spark 1.0.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/.gitignore +37 -0
- data/Gemfile +47 -0
- data/Guardfile +5 -0
- data/LICENSE.txt +22 -0
- data/README.md +185 -0
- data/Rakefile +35 -0
- data/TODO.md +7 -0
- data/benchmark/aggregate.rb +33 -0
- data/benchmark/bisect.rb +88 -0
- data/benchmark/custom_marshal.rb +94 -0
- data/benchmark/digest.rb +150 -0
- data/benchmark/enumerator.rb +88 -0
- data/benchmark/performance/prepare.sh +18 -0
- data/benchmark/performance/python.py +156 -0
- data/benchmark/performance/r.r +69 -0
- data/benchmark/performance/ruby.rb +167 -0
- data/benchmark/performance/run-all.sh +160 -0
- data/benchmark/performance/scala.scala +181 -0
- data/benchmark/serializer.rb +82 -0
- data/benchmark/sort.rb +43 -0
- data/benchmark/sort2.rb +164 -0
- data/benchmark/take.rb +28 -0
- data/bin/ruby-spark +8 -0
- data/example/pi.rb +28 -0
- data/ext/ruby_c/extconf.rb +3 -0
- data/ext/ruby_c/murmur.c +158 -0
- data/ext/ruby_c/murmur.h +9 -0
- data/ext/ruby_c/ruby-spark.c +18 -0
- data/ext/ruby_java/Digest.java +36 -0
- data/ext/ruby_java/Murmur2.java +98 -0
- data/ext/ruby_java/RubySparkExtService.java +28 -0
- data/ext/ruby_java/extconf.rb +3 -0
- data/ext/spark/build.sbt +73 -0
- data/ext/spark/project/plugins.sbt +9 -0
- data/ext/spark/sbt/sbt +34 -0
- data/ext/spark/src/main/scala/Exec.scala +91 -0
- data/ext/spark/src/main/scala/MLLibAPI.scala +4 -0
- data/ext/spark/src/main/scala/Marshal.scala +52 -0
- data/ext/spark/src/main/scala/MarshalDump.scala +113 -0
- data/ext/spark/src/main/scala/MarshalLoad.scala +220 -0
- data/ext/spark/src/main/scala/RubyAccumulatorParam.scala +69 -0
- data/ext/spark/src/main/scala/RubyBroadcast.scala +13 -0
- data/ext/spark/src/main/scala/RubyConstant.scala +13 -0
- data/ext/spark/src/main/scala/RubyMLLibAPI.scala +55 -0
- data/ext/spark/src/main/scala/RubyMLLibUtilAPI.scala +21 -0
- data/ext/spark/src/main/scala/RubyPage.scala +34 -0
- data/ext/spark/src/main/scala/RubyRDD.scala +364 -0
- data/ext/spark/src/main/scala/RubySerializer.scala +14 -0
- data/ext/spark/src/main/scala/RubyTab.scala +11 -0
- data/ext/spark/src/main/scala/RubyUtils.scala +15 -0
- data/ext/spark/src/main/scala/RubyWorker.scala +257 -0
- data/ext/spark/src/test/scala/MarshalSpec.scala +84 -0
- data/lib/ruby-spark.rb +1 -0
- data/lib/spark.rb +198 -0
- data/lib/spark/accumulator.rb +260 -0
- data/lib/spark/broadcast.rb +98 -0
- data/lib/spark/build.rb +43 -0
- data/lib/spark/cli.rb +169 -0
- data/lib/spark/command.rb +86 -0
- data/lib/spark/command/base.rb +154 -0
- data/lib/spark/command/basic.rb +345 -0
- data/lib/spark/command/pair.rb +124 -0
- data/lib/spark/command/sort.rb +51 -0
- data/lib/spark/command/statistic.rb +144 -0
- data/lib/spark/command_builder.rb +141 -0
- data/lib/spark/command_validator.rb +34 -0
- data/lib/spark/config.rb +244 -0
- data/lib/spark/constant.rb +14 -0
- data/lib/spark/context.rb +304 -0
- data/lib/spark/error.rb +50 -0
- data/lib/spark/ext/hash.rb +41 -0
- data/lib/spark/ext/integer.rb +25 -0
- data/lib/spark/ext/io.rb +57 -0
- data/lib/spark/ext/ip_socket.rb +29 -0
- data/lib/spark/ext/module.rb +58 -0
- data/lib/spark/ext/object.rb +24 -0
- data/lib/spark/ext/string.rb +24 -0
- data/lib/spark/helper.rb +10 -0
- data/lib/spark/helper/logger.rb +40 -0
- data/lib/spark/helper/parser.rb +85 -0
- data/lib/spark/helper/serialize.rb +71 -0
- data/lib/spark/helper/statistic.rb +93 -0
- data/lib/spark/helper/system.rb +42 -0
- data/lib/spark/java_bridge.rb +19 -0
- data/lib/spark/java_bridge/base.rb +203 -0
- data/lib/spark/java_bridge/jruby.rb +23 -0
- data/lib/spark/java_bridge/rjb.rb +41 -0
- data/lib/spark/logger.rb +76 -0
- data/lib/spark/mllib.rb +100 -0
- data/lib/spark/mllib/classification/common.rb +31 -0
- data/lib/spark/mllib/classification/logistic_regression.rb +223 -0
- data/lib/spark/mllib/classification/naive_bayes.rb +97 -0
- data/lib/spark/mllib/classification/svm.rb +135 -0
- data/lib/spark/mllib/clustering/gaussian_mixture.rb +82 -0
- data/lib/spark/mllib/clustering/kmeans.rb +118 -0
- data/lib/spark/mllib/matrix.rb +120 -0
- data/lib/spark/mllib/regression/common.rb +73 -0
- data/lib/spark/mllib/regression/labeled_point.rb +41 -0
- data/lib/spark/mllib/regression/lasso.rb +100 -0
- data/lib/spark/mllib/regression/linear.rb +124 -0
- data/lib/spark/mllib/regression/ridge.rb +97 -0
- data/lib/spark/mllib/ruby_matrix/matrix_adapter.rb +53 -0
- data/lib/spark/mllib/ruby_matrix/vector_adapter.rb +57 -0
- data/lib/spark/mllib/stat/distribution.rb +12 -0
- data/lib/spark/mllib/vector.rb +185 -0
- data/lib/spark/rdd.rb +1328 -0
- data/lib/spark/sampler.rb +92 -0
- data/lib/spark/serializer.rb +24 -0
- data/lib/spark/serializer/base.rb +170 -0
- data/lib/spark/serializer/cartesian.rb +37 -0
- data/lib/spark/serializer/marshal.rb +19 -0
- data/lib/spark/serializer/message_pack.rb +25 -0
- data/lib/spark/serializer/oj.rb +25 -0
- data/lib/spark/serializer/pair.rb +27 -0
- data/lib/spark/serializer/utf8.rb +25 -0
- data/lib/spark/sort.rb +189 -0
- data/lib/spark/stat_counter.rb +125 -0
- data/lib/spark/storage_level.rb +39 -0
- data/lib/spark/version.rb +3 -0
- data/lib/spark/worker/master.rb +144 -0
- data/lib/spark/worker/spark_files.rb +15 -0
- data/lib/spark/worker/worker.rb +197 -0
- data/ruby-spark.gemspec +36 -0
- data/spec/generator.rb +37 -0
- data/spec/inputs/lorem_300.txt +316 -0
- data/spec/inputs/numbers/1.txt +50 -0
- data/spec/inputs/numbers/10.txt +50 -0
- data/spec/inputs/numbers/11.txt +50 -0
- data/spec/inputs/numbers/12.txt +50 -0
- data/spec/inputs/numbers/13.txt +50 -0
- data/spec/inputs/numbers/14.txt +50 -0
- data/spec/inputs/numbers/15.txt +50 -0
- data/spec/inputs/numbers/16.txt +50 -0
- data/spec/inputs/numbers/17.txt +50 -0
- data/spec/inputs/numbers/18.txt +50 -0
- data/spec/inputs/numbers/19.txt +50 -0
- data/spec/inputs/numbers/2.txt +50 -0
- data/spec/inputs/numbers/20.txt +50 -0
- data/spec/inputs/numbers/3.txt +50 -0
- data/spec/inputs/numbers/4.txt +50 -0
- data/spec/inputs/numbers/5.txt +50 -0
- data/spec/inputs/numbers/6.txt +50 -0
- data/spec/inputs/numbers/7.txt +50 -0
- data/spec/inputs/numbers/8.txt +50 -0
- data/spec/inputs/numbers/9.txt +50 -0
- data/spec/inputs/numbers_0_100.txt +101 -0
- data/spec/inputs/numbers_1_100.txt +100 -0
- data/spec/lib/collect_spec.rb +42 -0
- data/spec/lib/command_spec.rb +68 -0
- data/spec/lib/config_spec.rb +64 -0
- data/spec/lib/context_spec.rb +163 -0
- data/spec/lib/ext_spec.rb +72 -0
- data/spec/lib/external_apps_spec.rb +45 -0
- data/spec/lib/filter_spec.rb +80 -0
- data/spec/lib/flat_map_spec.rb +100 -0
- data/spec/lib/group_spec.rb +109 -0
- data/spec/lib/helper_spec.rb +19 -0
- data/spec/lib/key_spec.rb +41 -0
- data/spec/lib/manipulation_spec.rb +114 -0
- data/spec/lib/map_partitions_spec.rb +87 -0
- data/spec/lib/map_spec.rb +91 -0
- data/spec/lib/mllib/classification_spec.rb +54 -0
- data/spec/lib/mllib/clustering_spec.rb +35 -0
- data/spec/lib/mllib/matrix_spec.rb +32 -0
- data/spec/lib/mllib/regression_spec.rb +116 -0
- data/spec/lib/mllib/vector_spec.rb +77 -0
- data/spec/lib/reduce_by_key_spec.rb +118 -0
- data/spec/lib/reduce_spec.rb +131 -0
- data/spec/lib/sample_spec.rb +46 -0
- data/spec/lib/serializer_spec.rb +13 -0
- data/spec/lib/sort_spec.rb +58 -0
- data/spec/lib/statistic_spec.rb +168 -0
- data/spec/lib/whole_text_files_spec.rb +33 -0
- data/spec/spec_helper.rb +39 -0
- metadata +301 -0
@@ -0,0 +1,109 @@
|
|
1
|
+
require "spec_helper"
|
2
|
+
|
3
|
+
RSpec::shared_examples "a groupping by key" do |workers|
|
4
|
+
it "with #{workers || 'default'} worker" do
|
5
|
+
expect(rdd_result(workers)).to eql(result)
|
6
|
+
end
|
7
|
+
end
|
8
|
+
|
9
|
+
RSpec::shared_examples "a cogroupping by key" do |workers|
|
10
|
+
context "with #{workers || 'default'} worker" do
|
11
|
+
it ".group_with" do
|
12
|
+
rdd = rdd_1(workers).group_with(rdd_2(workers))
|
13
|
+
expect(rdd.collect_as_hash).to eql(result_12)
|
14
|
+
end
|
15
|
+
|
16
|
+
it ".cogroup" do
|
17
|
+
rdd = rdd_1(workers).cogroup(rdd_2(workers), rdd_3(workers))
|
18
|
+
expect(rdd.collect_as_hash).to eql(result_123)
|
19
|
+
end
|
20
|
+
end
|
21
|
+
end
|
22
|
+
|
23
|
+
RSpec::shared_examples "a groupping by" do |workers|
|
24
|
+
it "with #{workers || 'default'} worker" do
|
25
|
+
rdd = rdd_numbers(workers)
|
26
|
+
rdd = rdd.group_by(key_function1)
|
27
|
+
|
28
|
+
expect(rdd.collect_as_hash).to eql(numbers.group_by(&key_function1))
|
29
|
+
|
30
|
+
rdd = rdd_words(workers)
|
31
|
+
rdd = rdd.group_by(key_function2)
|
32
|
+
|
33
|
+
expect(rdd.collect_as_hash).to eql(words.group_by(&key_function2))
|
34
|
+
end
|
35
|
+
end
|
36
|
+
|
37
|
+
RSpec::describe "Spark::RDD" do
|
38
|
+
|
39
|
+
def make_result(*hashes)
|
40
|
+
_result = {}
|
41
|
+
hashes.each do |data|
|
42
|
+
data.each do |key, value|
|
43
|
+
_result[key] ||= []
|
44
|
+
_result[key] << value
|
45
|
+
end
|
46
|
+
end
|
47
|
+
_result
|
48
|
+
end
|
49
|
+
|
50
|
+
context ".group_by_key" do
|
51
|
+
let(:hash) { Generator.hash }
|
52
|
+
let(:result) { make_result(hash) }
|
53
|
+
|
54
|
+
def rdd_result(workers)
|
55
|
+
rdd = $sc.parallelize(hash)
|
56
|
+
rdd.group_by_key.collect_as_hash
|
57
|
+
end
|
58
|
+
|
59
|
+
it_behaves_like "a groupping by key", nil
|
60
|
+
it_behaves_like "a groupping by key", 1
|
61
|
+
it_behaves_like "a groupping by key", rand(2..10)
|
62
|
+
end
|
63
|
+
|
64
|
+
context "cogroup" do
|
65
|
+
let(:hash1) { Generator.hash }
|
66
|
+
let(:hash2) { Generator.hash }
|
67
|
+
let(:hash3) { Generator.hash }
|
68
|
+
|
69
|
+
let(:result_12) { make_result(hash1, hash2) }
|
70
|
+
let(:result_123) { make_result(hash1, hash2, hash3) }
|
71
|
+
|
72
|
+
def rdd_1(workers)
|
73
|
+
$sc.parallelize(hash1)
|
74
|
+
end
|
75
|
+
|
76
|
+
def rdd_2(workers)
|
77
|
+
$sc.parallelize(hash2)
|
78
|
+
end
|
79
|
+
|
80
|
+
def rdd_3(workers)
|
81
|
+
$sc.parallelize(hash3)
|
82
|
+
end
|
83
|
+
|
84
|
+
it_behaves_like "a cogroupping by key", nil
|
85
|
+
it_behaves_like "a cogroupping by key", 1
|
86
|
+
it_behaves_like "a cogroupping by key", rand(2..10)
|
87
|
+
end
|
88
|
+
|
89
|
+
context "group_by" do
|
90
|
+
let(:key_function1) { lambda{|x| x%2} }
|
91
|
+
let(:key_function2) { lambda{|x| x.size} }
|
92
|
+
|
93
|
+
let(:numbers) { Generator.numbers }
|
94
|
+
let(:words) { Generator.words }
|
95
|
+
|
96
|
+
def rdd_numbers(workers)
|
97
|
+
$sc.parallelize(numbers)
|
98
|
+
end
|
99
|
+
|
100
|
+
def rdd_words(workers)
|
101
|
+
$sc.parallelize(words)
|
102
|
+
end
|
103
|
+
|
104
|
+
it_behaves_like "a groupping by", nil
|
105
|
+
it_behaves_like "a groupping by", 1
|
106
|
+
it_behaves_like "a groupping by", rand(2..10)
|
107
|
+
end
|
108
|
+
|
109
|
+
end
|
@@ -0,0 +1,19 @@
|
|
1
|
+
require "spec_helper"
|
2
|
+
|
3
|
+
RSpec::configure do |c|
|
4
|
+
c.include Spark::Helper::Parser
|
5
|
+
end
|
6
|
+
|
7
|
+
RSpec::describe Spark::Helper do
|
8
|
+
|
9
|
+
it "memory size" do
|
10
|
+
expect(to_memory_size("512mb")).to eql(524288.0)
|
11
|
+
expect(to_memory_size("1586 mb")).to eql(1624064.0)
|
12
|
+
expect(to_memory_size("3 MB")).to eql(3072.0)
|
13
|
+
expect(to_memory_size("9gb")).to eql(9437184.0)
|
14
|
+
expect(to_memory_size("9gb", "mb")).to eql(9216.0)
|
15
|
+
expect(to_memory_size("9mb", "gb")).to eql(0.01)
|
16
|
+
expect(to_memory_size("6652548796kb", "mb")).to eql(6496629.68)
|
17
|
+
end
|
18
|
+
|
19
|
+
end
|
@@ -0,0 +1,41 @@
|
|
1
|
+
require "spec_helper"
|
2
|
+
|
3
|
+
RSpec::shared_examples "a keying by" do |workers|
|
4
|
+
it "with #{workers || 'default'} worker" do
|
5
|
+
rdd = rdd_numbers(workers)
|
6
|
+
rdd = rdd.key_by(key_function1)
|
7
|
+
|
8
|
+
result = numbers.map{|item| [key_function1.call(item), item]}
|
9
|
+
expect(rdd.collect).to eql(result)
|
10
|
+
|
11
|
+
rdd = rdd_words(workers)
|
12
|
+
rdd = rdd.key_by(key_function2)
|
13
|
+
|
14
|
+
result = words.map{|item| [key_function2.call(item), item]}
|
15
|
+
expect(rdd.collect).to eql(result)
|
16
|
+
end
|
17
|
+
end
|
18
|
+
|
19
|
+
RSpec::describe "Spark::RDD" do
|
20
|
+
|
21
|
+
context "key_by" do
|
22
|
+
let(:key_function1) { lambda{|x| x.even?} }
|
23
|
+
let(:key_function2) { lambda{|x| x.include?("a")} }
|
24
|
+
|
25
|
+
let(:numbers) { Generator.numbers }
|
26
|
+
let(:words) { Generator.words }
|
27
|
+
|
28
|
+
def rdd_numbers(workers)
|
29
|
+
$sc.parallelize(numbers)
|
30
|
+
end
|
31
|
+
|
32
|
+
def rdd_words(workers)
|
33
|
+
$sc.parallelize(words)
|
34
|
+
end
|
35
|
+
|
36
|
+
it_behaves_like "a keying by", nil
|
37
|
+
it_behaves_like "a keying by", 1
|
38
|
+
it_behaves_like "a keying by", rand(2..10)
|
39
|
+
end
|
40
|
+
|
41
|
+
end
|
@@ -0,0 +1,114 @@
|
|
1
|
+
require "spec_helper"
|
2
|
+
|
3
|
+
RSpec::describe "Spark::RDD" do
|
4
|
+
let(:numbers) { 1..100 }
|
5
|
+
let(:rand_numbers) { Generator.numbers }
|
6
|
+
|
7
|
+
it ".glom" do
|
8
|
+
rdd = $sc.parallelize(numbers, 1).glom
|
9
|
+
expect(rdd.collect).to eql([numbers.to_a])
|
10
|
+
|
11
|
+
rdd = $sc.parallelize(numbers, 5, batch_size: 1).glom
|
12
|
+
expect(rdd.collect).to eql(numbers.each_slice(20).to_a)
|
13
|
+
end
|
14
|
+
|
15
|
+
it ".coalesce" do
|
16
|
+
rdd = $sc.parallelize(numbers, 5)
|
17
|
+
|
18
|
+
rdd2 = rdd.glom
|
19
|
+
expect(rdd2.collect.size).to eql(5)
|
20
|
+
|
21
|
+
rdd3 = rdd.coalesce(4).glom
|
22
|
+
expect(rdd3.collect.size).to eql(4)
|
23
|
+
end
|
24
|
+
|
25
|
+
it ".distinct" do
|
26
|
+
rdd = $sc.parallelize(rand_numbers, 5)
|
27
|
+
rdd = rdd.distinct
|
28
|
+
expect(rdd.collect.sort).to eql(rand_numbers.uniq.sort)
|
29
|
+
|
30
|
+
rdd = $sc.parallelize(numbers, 5)
|
31
|
+
rdd = rdd.map(lambda{|x| 1})
|
32
|
+
rdd = rdd.distinct
|
33
|
+
expect(rdd.collect).to eql([1])
|
34
|
+
end
|
35
|
+
|
36
|
+
context ".union" do
|
37
|
+
it "classic method" do
|
38
|
+
rdd = $sc.parallelize(numbers, 5)
|
39
|
+
rdd = rdd.union(rdd).collect
|
40
|
+
|
41
|
+
expect(rdd.collect.sort).to eql((numbers.to_a+numbers.to_a).sort)
|
42
|
+
end
|
43
|
+
|
44
|
+
it "with a different serializer" do
|
45
|
+
rdd1 = $sc.parallelize(numbers, 1, serializer: "marshal")
|
46
|
+
rdd2 = $sc.parallelize(numbers, 1, serializer: "oj")
|
47
|
+
|
48
|
+
expect { rdd1.union(rdd2).collect }.to_not raise_error
|
49
|
+
end
|
50
|
+
|
51
|
+
it "as operator" do
|
52
|
+
rdd1 = $sc.parallelize(numbers)
|
53
|
+
rdd2 = $sc.parallelize(rand_numbers)
|
54
|
+
|
55
|
+
expect((rdd1+rdd2).sum).to eql((numbers.to_a+rand_numbers).reduce(:+))
|
56
|
+
end
|
57
|
+
end
|
58
|
+
|
59
|
+
it ".compact" do
|
60
|
+
data = [nil, nil , 0, 0, 1, 2, nil, 6]
|
61
|
+
result = data.compact
|
62
|
+
|
63
|
+
rdd = $sc.parallelize(data, 1).compact
|
64
|
+
expect(rdd.collect).to eql(result)
|
65
|
+
|
66
|
+
rdd = $sc.parallelize(data, 5, batch_size: 1).compact
|
67
|
+
expect(rdd.collect).to eql(result)
|
68
|
+
|
69
|
+
rdd = $sc.parallelize(data, 1, batch_size: 1).compact
|
70
|
+
expect(rdd.collect).to eql(result)
|
71
|
+
end
|
72
|
+
|
73
|
+
it ".intersection" do
|
74
|
+
data1 = [0,1,2,3,4,5,6,7,8,9,10]
|
75
|
+
data2 = [5,6,7,8,9,10,11,12,13,14,15]
|
76
|
+
|
77
|
+
rdd1 = $sc.parallelize(data1)
|
78
|
+
rdd2 = $sc.parallelize(data2)
|
79
|
+
|
80
|
+
expect(rdd1.intersection(rdd2).collect.sort).to eql(data1 & data2)
|
81
|
+
end
|
82
|
+
|
83
|
+
it ".shuffle" do
|
84
|
+
data = Generator.numbers
|
85
|
+
rdd = $sc.parallelize(data)
|
86
|
+
|
87
|
+
expect(rdd.shuffle.collect).to_not eql(data)
|
88
|
+
end
|
89
|
+
|
90
|
+
context ".cartesian" do
|
91
|
+
let(:data1) { Generator.numbers(100) }
|
92
|
+
let(:data2) { Generator.numbers(100) }
|
93
|
+
let(:result) { data1.product(data2).map(&:to_s).sort }
|
94
|
+
|
95
|
+
it "unbatched" do
|
96
|
+
rdd1 = $sc.parallelize(data1, 2, batch_size: 1)
|
97
|
+
rdd2 = $sc.parallelize(data2, 2, batch_size: 1)
|
98
|
+
|
99
|
+
rdd = rdd1.cartesian(rdd2).map(lambda{|x| x.to_s})
|
100
|
+
|
101
|
+
expect(rdd.collect.sort).to eql(result)
|
102
|
+
end
|
103
|
+
|
104
|
+
it "batched" do
|
105
|
+
rdd1 = $sc.parallelize(data1, 2, batch_size: rand(4..10))
|
106
|
+
rdd2 = $sc.parallelize(data2, 2, batch_size: rand(4..10))
|
107
|
+
|
108
|
+
rdd = rdd1.cartesian(rdd2).map(lambda{|x| x.to_s})
|
109
|
+
|
110
|
+
expect(rdd.collect.sort).to eql(result)
|
111
|
+
end
|
112
|
+
end
|
113
|
+
|
114
|
+
end
|
@@ -0,0 +1,87 @@
|
|
1
|
+
require "spec_helper"
|
2
|
+
|
3
|
+
def func3(x)
|
4
|
+
x.map(&:to_i).reduce(:+)
|
5
|
+
end
|
6
|
+
|
7
|
+
def func4_with_index(data, index)
|
8
|
+
{
|
9
|
+
index => data.map(&:to_i).reduce(:*)
|
10
|
+
}
|
11
|
+
end
|
12
|
+
|
13
|
+
RSpec::shared_examples "a map partitions" do |workers|
|
14
|
+
context "with #{workers || 'default'} worker" do
|
15
|
+
it "without index" do
|
16
|
+
rdd2 = rdd(workers).map_partitions(func1)
|
17
|
+
result = func1.call(numbers)
|
18
|
+
|
19
|
+
expect(func1.call(rdd2.collect)).to eql(result)
|
20
|
+
|
21
|
+
rdd3 = rdd(workers)
|
22
|
+
rdd3 = rdd3.map_partitions(func1)
|
23
|
+
rdd3 = rdd3.map_partitions(func2)
|
24
|
+
rdd3 = rdd3.map_partitions(method(:func3))
|
25
|
+
result = func3(func2.call(func1.call(numbers)))
|
26
|
+
|
27
|
+
# Not same number of workers
|
28
|
+
expect(rdd3.collect.size).to be >= 1
|
29
|
+
|
30
|
+
rdd4 = rdd(workers)
|
31
|
+
rdd4 = rdd4.map_partitions(func1)
|
32
|
+
rdd4 = rdd4.map_partitions(func2)
|
33
|
+
rdd4 = rdd4.map_partitions(method(:func3))
|
34
|
+
|
35
|
+
expect(rdd4.collect).to eql(rdd3.collect)
|
36
|
+
end
|
37
|
+
|
38
|
+
it "with index" do
|
39
|
+
rdd2 = rdd(workers).map_partitions_with_index(method(:func4_with_index))
|
40
|
+
result = rdd2.collect
|
41
|
+
|
42
|
+
expect(result).to be_a(Array)
|
43
|
+
|
44
|
+
result.each do |x|
|
45
|
+
expect(x).to be_a(Hash)
|
46
|
+
end
|
47
|
+
|
48
|
+
# Multiply by 0
|
49
|
+
# Some values are 0 because of batched serialization
|
50
|
+
expect(result.map(&:values).flatten.compact.uniq.first).to eql(0)
|
51
|
+
end
|
52
|
+
end
|
53
|
+
end
|
54
|
+
|
55
|
+
RSpec::describe "Spark::RDD.map_partitions(_with_index)" do
|
56
|
+
let(:func1) { lambda{|x| x.map(&:to_i)} }
|
57
|
+
let(:func2) {
|
58
|
+
lambda{|x|
|
59
|
+
x.map{|y| y*2}
|
60
|
+
}
|
61
|
+
}
|
62
|
+
|
63
|
+
context "throught parallelize" do
|
64
|
+
let(:numbers) { 0..1000 }
|
65
|
+
|
66
|
+
def rdd(workers)
|
67
|
+
$sc.parallelize(numbers, workers)
|
68
|
+
end
|
69
|
+
|
70
|
+
it_behaves_like "a map partitions", nil
|
71
|
+
it_behaves_like "a map partitions", 1
|
72
|
+
it_behaves_like "a map partitions", rand(2..10)
|
73
|
+
end
|
74
|
+
|
75
|
+
context "throught text_file" do
|
76
|
+
let(:file) { File.join("spec", "inputs", "numbers_0_100.txt") }
|
77
|
+
let(:numbers) { File.readlines(file).map(&:strip) }
|
78
|
+
|
79
|
+
def rdd(workers)
|
80
|
+
$sc.text_file(file, workers)
|
81
|
+
end
|
82
|
+
|
83
|
+
it_behaves_like "a map partitions", nil
|
84
|
+
it_behaves_like "a map partitions", 1
|
85
|
+
it_behaves_like "a map partitions", rand(2..10)
|
86
|
+
end
|
87
|
+
end
|
@@ -0,0 +1,91 @@
|
|
1
|
+
require "spec_helper"
|
2
|
+
|
3
|
+
RSpec::shared_examples "a mapping" do |workers|
|
4
|
+
it "with #{workers || 'default'} worker" do
|
5
|
+
rdd2 = rdd(workers).map(func1)
|
6
|
+
result = numbers.map(&func1)
|
7
|
+
|
8
|
+
expect(rdd2.collect).to eql(result)
|
9
|
+
|
10
|
+
rdd3 = rdd(workers)
|
11
|
+
rdd3 = rdd3.map(func1)
|
12
|
+
rdd3 = rdd3.map(func2)
|
13
|
+
rdd3 = rdd3.map(func3)
|
14
|
+
result = numbers.map(&func1).map(&func2).map(&func3)
|
15
|
+
|
16
|
+
expect(rdd3.collect).to eql(result)
|
17
|
+
|
18
|
+
rdd4 = rdd(workers)
|
19
|
+
rdd4 = rdd4.map(func3)
|
20
|
+
rdd4 = rdd4.map(func2)
|
21
|
+
rdd4 = rdd4.map(func1)
|
22
|
+
|
23
|
+
expect(rdd4.collect).to eql(rdd3.collect)
|
24
|
+
end
|
25
|
+
end
|
26
|
+
|
27
|
+
RSpec::shared_examples "a mapping values" do |workers|
|
28
|
+
it "with #{workers || 'default'} worker" do
|
29
|
+
rdd2 = rdd(workers).map_values(func1)
|
30
|
+
result = hash.map{|key, value| [key, func1.call(value)]}
|
31
|
+
|
32
|
+
expect(rdd2.collect).to eql(result)
|
33
|
+
|
34
|
+
rdd3 = rdd(workers)
|
35
|
+
rdd3 = rdd3.map_values(func1)
|
36
|
+
rdd3 = rdd3.map_values(func2)
|
37
|
+
rdd3 = rdd3.map_values(func3)
|
38
|
+
result = hash.map{|key, value| [key, func1.call(value)]}
|
39
|
+
.map{|key, value| [key, func2.call(value)]}
|
40
|
+
.map{|key, value| [key, func3.call(value)]}
|
41
|
+
|
42
|
+
expect(rdd3.collect).to eql(result)
|
43
|
+
end
|
44
|
+
end
|
45
|
+
|
46
|
+
RSpec::describe "Spark::RDD" do
|
47
|
+
let(:func1) { lambda{|x| x*2} }
|
48
|
+
let(:func2) { lambda{|x| x*3} }
|
49
|
+
let(:func3) { lambda{|x| x*4} }
|
50
|
+
|
51
|
+
context "throught parallelize" do
|
52
|
+
context ".map" do
|
53
|
+
let(:numbers) { Generator.numbers }
|
54
|
+
|
55
|
+
def rdd(workers)
|
56
|
+
$sc.parallelize(numbers, workers)
|
57
|
+
end
|
58
|
+
|
59
|
+
it_behaves_like "a mapping", nil
|
60
|
+
it_behaves_like "a mapping", 1
|
61
|
+
it_behaves_like "a mapping", rand(2..10)
|
62
|
+
end
|
63
|
+
|
64
|
+
context ".map_values" do
|
65
|
+
let!(:hash) { Generator.hash }
|
66
|
+
|
67
|
+
def rdd(workers)
|
68
|
+
$sc.parallelize(hash, workers)
|
69
|
+
end
|
70
|
+
|
71
|
+
it_behaves_like "a mapping values", nil
|
72
|
+
it_behaves_like "a mapping values", 1
|
73
|
+
it_behaves_like "a mapping values", rand(2..10)
|
74
|
+
end
|
75
|
+
end
|
76
|
+
|
77
|
+
context "throught text_file" do
|
78
|
+
context ".map" do
|
79
|
+
let(:file) { File.join("spec", "inputs", "numbers_0_100.txt") }
|
80
|
+
let(:numbers) { File.readlines(file).map(&:strip) }
|
81
|
+
|
82
|
+
def rdd(workers)
|
83
|
+
$sc.text_file(file, workers)
|
84
|
+
end
|
85
|
+
|
86
|
+
it_behaves_like "a mapping", nil
|
87
|
+
it_behaves_like "a mapping", 1
|
88
|
+
it_behaves_like "a mapping", rand(2..10)
|
89
|
+
end
|
90
|
+
end
|
91
|
+
end
|