Merge pull request #53 from DataSketches/JaccardSimilarityUDF
Jaccard similarity udf
diff --git a/src/main/java/com/yahoo/sketches/pig/quantiles/DataToDoublesSketch.java b/src/main/java/com/yahoo/sketches/pig/quantiles/DataToDoublesSketch.java
index 84ca5d9..d1bc8df 100644
--- a/src/main/java/com/yahoo/sketches/pig/quantiles/DataToDoublesSketch.java
+++ b/src/main/java/com/yahoo/sketches/pig/quantiles/DataToDoublesSketch.java
@@ -124,12 +124,12 @@
@Override // TOP LEVEL EXEC
public Tuple exec(final Tuple inputTuple) throws IOException {
//The exec is a stateless function. It operates on the input and returns a result.
- if (inputTuple != null && inputTuple.size() > 0) {
+ if ((inputTuple != null) && (inputTuple.size() > 0)) {
final DoublesUnion union = unionBuilder_.build();
final DataBag bag = (DataBag) inputTuple.get(0);
for (final Tuple innerTuple: bag) {
- Object value = innerTuple.get(0);
- if(value != null) {
+ final Object value = innerTuple.get(0);
+ if (value != null) {
union.update((Double) value);
}
}
@@ -172,15 +172,15 @@
*/
@Override
public void accumulate(final Tuple inputTuple) throws IOException {
- if (inputTuple == null || inputTuple.size() == 0) { return; }
+ if ((inputTuple == null) || (inputTuple.size() == 0)) { return; }
final DataBag bag = (DataBag) inputTuple.get(0);
if (bag == null) { return; }
if (accumUnion_ == null) {
accumUnion_ = unionBuilder_.build();
}
for (final Tuple innerTuple: bag) {
- Object value = innerTuple.get(0);
- if(value != null) {
+ final Object value = innerTuple.get(0);
+ if (value != null) {
accumUnion_.update((Double) value);
}
}
@@ -309,7 +309,7 @@
@Override // IntermediateFinal exec
public Tuple exec(final Tuple inputTuple) throws IOException { //throws is in API
- if (inputTuple != null && inputTuple.size() > 0) {
+ if ((inputTuple != null) && (inputTuple.size() > 0)) {
final DoublesUnion union = unionBuilder_.build();
final DataBag outerBag = (DataBag) inputTuple.get(0);
for (final Tuple dataTuple: outerBag) {
@@ -323,8 +323,8 @@
// It is due to system bagged outputs from multiple mapper Initial functions.
// The Intermediate stage was bypassed.
for (final Tuple innerTuple: innerBag) {
- Object value = innerTuple.get(0);
- if(value != null) {
+ final Object value = innerTuple.get(0);
+ if (value != null) {
union.update((Double) value);
}
}
diff --git a/src/main/java/com/yahoo/sketches/pig/quantiles/DataToItemsSketch.java b/src/main/java/com/yahoo/sketches/pig/quantiles/DataToItemsSketch.java
index c1d28bb..fcccfb6 100644
--- a/src/main/java/com/yahoo/sketches/pig/quantiles/DataToItemsSketch.java
+++ b/src/main/java/com/yahoo/sketches/pig/quantiles/DataToItemsSketch.java
@@ -116,14 +116,14 @@
@Override // TOP LEVEL EXEC
public Tuple exec(final Tuple inputTuple) throws IOException {
//The exec is a stateless function. It operates on the input and returns a result.
- if (inputTuple != null && inputTuple.size() > 0) {
+ if ((inputTuple != null) && (inputTuple.size() > 0)) {
final ItemsUnion<T> union = k_ > 0
? ItemsUnion.getInstance(k_, comparator_)
: ItemsUnion.getInstance(comparator_);
final DataBag bag = (DataBag) inputTuple.get(0);
for (final Tuple innerTuple: bag) {
- Object value = innerTuple.get(0);
- if(value != null) {
+ final Object value = innerTuple.get(0);
+ if (value != null) {
union.update(extractValue(value));
}
}
@@ -167,7 +167,7 @@
*/
@Override
public void accumulate(final Tuple inputTuple) throws IOException {
- if (inputTuple == null || inputTuple.size() == 0) { return; }
+ if ((inputTuple == null) || (inputTuple.size() == 0)) { return; }
final DataBag bag = (DataBag) inputTuple.get(0);
if (bag == null) { return; }
if (accumUnion_ == null) {
@@ -176,8 +176,8 @@
: ItemsUnion.getInstance(comparator_);
}
for (final Tuple innerTuple: bag) {
- Object value = innerTuple.get(0);
- if(value != null) {
+ final Object value = innerTuple.get(0);
+ if (value != null) {
accumUnion_.update(extractValue(value));
}
}
@@ -297,7 +297,7 @@
@Override // IntermediateFinal exec
public Tuple exec(final Tuple inputTuple) throws IOException { //throws is in API
- if (inputTuple != null && inputTuple.size() > 0) {
+ if ((inputTuple != null) && (inputTuple.size() > 0)) {
final ItemsUnion<T> union = k_ > 0
? ItemsUnion.getInstance(k_, comparator_)
: ItemsUnion.getInstance(comparator_);
@@ -313,8 +313,8 @@
// It is due to system bagged outputs from multiple mapper Initial functions.
// The Intermediate stage was bypassed.
for (final Tuple innerTuple: innerBag) {
- Object value = innerTuple.get(0);
- if(value != null) {
+ final Object value = innerTuple.get(0);
+ if (value != null) {
union.update(extractValue(value));
}
}
diff --git a/tools/SketchesCheckstyle.xml b/tools/SketchesCheckstyle.xml
index bfdf482..0a67506 100644
--- a/tools/SketchesCheckstyle.xml
+++ b/tools/SketchesCheckstyle.xml
@@ -4,34 +4,21 @@
"http://www.puppycrawl.com/dtds/configuration_1_3.dtd"> <!-- does not work with https -->
<!--
- This configuration file was written by the eclipse-cs plugin configuration editor
- Checkstyle-Configuration: SketchesCheckstyle
+ SketchesCheckstyle.xml for sketches-core
Checkstyle is very configurable. Be sure to read the documentation at
http://checkstyle.sourceforge.net (or in your downloaded distribution). Note: Does not work with https.
- Most Checks are configurable, be sure to consult the documentation.
-
To completely disable a check, just comment it out or delete it from the file.
Authors: Max Vetrenko, Ruslan Diachenko, Roman Ivanov.
-
-->
<module name = "Checker">
<property name="charset" value="UTF-8"/>
-
<property name="severity" value="warning"/>
-
<property name="fileExtensions" value="java, properties, xml"/>
-
- <!-- Enable suppression using comments: //CHECKSTYLE.OFF RULE and //CHECKSTYLE.ON RULE-->
- <module name="SuppressionCommentFilter">
- <property name="offCommentFormat" value="CHECKSTYLE.OFF\: ([\w\|]+)"/>
- <property name="onCommentFormat" value="CHECKSTYLE.ON\: ([\w\|]+)"/>
- <property name="checkFormat" value="$1"/>
- </module>
-
+
<module name="FileTabCharacter">
<property name="eachLine" value="true"/>
</module>
@@ -53,7 +40,7 @@
<property name="tokens" value="VARIABLE_DEF"/>
<property name="allowSamelineMultipleAnnotations" value="true"/>
</module>
-
+
<!-- Block Checks -->
<module name="EmptyBlock">
<property name="severity" value="warning"/>
@@ -71,7 +58,6 @@
<module name="LeftCurly">
<!-- doesn't allow for if (n == 0) { return 0.0; }, which is readable and not corruptable -->
<property name="severity" value="ignore"/>
- <property name="maxLineLength" value="100"/>
<metadata name="net.sf.eclipsecs.core.lastEnabledSeverity" value="inherit"/>
</module>
@@ -144,7 +130,6 @@
<module name="FinalClass"/>
-
<!-- Imports -->
<module name="AvoidStarImport"/>
@@ -160,6 +145,15 @@
<module name="UnusedImports"/>
+ <!-- Filters -->
+ <!-- Enable suppression using comments: //CHECKSTYLE.OFF "RULE" and //CHECKSTYLE.ON "RULE"
+ You must specify the specific rule, as in: //CHECKSTYLE.OFF: LineLength -->
+ <module name="SuppressionCommentFilter">
+ <property name="offCommentFormat" value="CHECKSTYLE.OFF\: ([\w\|]+)"/>
+ <property name="onCommentFormat" value="CHECKSTYLE.ON\: ([\w\|]+)"/>
+ <property name="checkFormat" value="$1"/>
+ </module>
+
<!-- Javadoc Comments -->
<!-- JavadocPackage under Checker -->
<module name="AtclauseOrder">
@@ -215,8 +209,6 @@
<!-- <metadata name="net.sf.eclipsecs.core.lastEnabledSeverity" value="inherit"/> -->
</module>
- <module name="FileContentsHolder"/> <!-- Used with SuppressionCommentFilter -->
-
<module name="Indentation">
<property name="severity" value="ignore"/>
<property name="basicOffset" value="2"/>
