add implementation with Krangl library (Kotlin)

uwemaurer · uwemaurer · commit bd0e5a4385b2 · 2020-01-28T10:10:09.000+01:00
diff --git a/build.gradle b/build.gradle
@@ -1,4 +1,17 @@
+buildscript {
+  ext.kotlin_version = '1.3.61'
+  repositories {
+    mavenCentral()
+    jcenter()
+  }
+
+  dependencies {
+    classpath "org.jetbrains.kotlin:kotlin-gradle-plugin:$kotlin_version"
+  }
+}
+
 apply plugin: 'java'
+apply plugin: 'kotlin'
 apply plugin: 'eclipse'
 
 sourceCompatibility = 1.8
@@ -9,15 +22,18 @@ repositories {
 }
 
 dependencies {
-  compile 'tech.tablesaw:tablesaw-core:0.37.3'
+  implementation 'tech.tablesaw:tablesaw-core:0.37.3'
 
-  compile 'joinery:joinery-dataframe:1.9'
+  implementation 'joinery:joinery-dataframe:1.9'
   // For the CSV import joinery needs this dependency too:
-  compile 'org.apache.poi:poi:3.17'
+  implementation 'org.apache.poi:poi:3.17'
+
+  implementation 'com.zavtech:morpheus-core:0.9.21'
 
-  compile 'com.zavtech:morpheus-core:0.9.21'
+  implementation 'org.datavec:datavec-api:1.0.0-beta2'
+  implementation 'org.datavec:datavec-local:1.0.0-beta2'
 
-  compile 'org.datavec:datavec-api:1.0.0-beta2'
-  compile 'org.datavec:datavec-local:1.0.0-beta2'
+  implementation "de.mpicbg.scicomp:krangl:0.11"
+  implementation "org.jetbrains.kotlin:kotlin-stdlib-jdk8:$kotlin_version"
 }
 
diff --git a/src/main/java/test_dataframes/TestKrangl.kt b/src/main/java/test_dataframes/TestKrangl.kt
@@ -0,0 +1,39 @@
+package test_dataframes
+
+import com.google.common.base.Stopwatch
+import krangl.*
+
+/**
+ * Test the API of krangl to do some basic dataframe manipulations.
+ *
+ * https://github.com/holgerbrandl/krangl
+ *
+ * See https://medium.com/@thijser/doing-cool-data-science-in-java-how-3-dataframe-libraries-stack-up-5e6ccb7b437
+ * for more information.
+ */
+fun main() {
+    val data = DataFrame.readCSV("urb_cpop1_1_Data.csv")
+
+    val watch = Stopwatch.createStarted()
+    // remove missing values indicated with ":", convert column to IntCol
+    val filtered = data.filter { !(it["Value"] eq ":") }.addColumn("Value") {
+        it["Value"].map(String::toInt)
+    }
+    // replace duplicated rows with mean value, create pivot table
+    val cities = filtered.groupBy("CITIES", "INDIC_UR", "TIME")
+            .summarize("Value" to { it["Value"].mean() })
+            .spread("TIME", "Value").filter {
+                it["INDIC_UR"].isMatching<String> { endsWith("January, total") }
+            }
+
+    println(cities.select("CITIES", "2017").sortedByDescending("2017").head(10))
+
+    val highestGrowthTable = cities.addColumn("growth") {
+        (it["2016"] / it["2010"] - 1.0) * 100.0
+    }.sortedByDescending("growth")
+
+    println(highestGrowthTable.select("CITIES", "growth").head(10))
+
+    CheckResult.checkResult(highestGrowthTable["CITIES"].asStrings().toList())
+    println("Total time: $watch")
+}
