Add Highcharts support to Apache Spark. Both in spark-shell and Zeppelin.

Make Spark DataFrame visualization simple, flexible, and beautiful.

Install docker if you have not.

docker run -p 8080:8080 -d knockdata/zeppelin-highcharts

the docker image zeppelin-highcharts includes features

• highcharts
• pyspark, include
• git, git lfs, vim

0.6.0: Spark 1.6, Scala 2.10, Zeppelin 0.6.0 0.6.1: Spark 2.0, Scala 2.11, Zeppelin 0.6.1

If you wanna run on your existing zeppelin, follow Use In Zeppelin.

http://localhost:8080

• Open and execute Zeppelin Tutorial NoteBook

Add a sub praragraph in Zeppelin Tutorial NoteBook

Paste the following code and execute it.

import com.knockdata.spark.highcharts._
import com.knockdata.spark.highcharts.model._

highcharts(bank
  .series("x" -> "age", "y" -> avg(col("balance")))
  .orderBy(col("age"))).plot()

You will get the following graph

The default version before 0.6 support Spark 1.6. In order to use Spark 2.0, compile with

mvn clean package -Pspark-2.0 -DskipTests

The artifact in the spark interpreter need use jar file for spark-highcharts. And the lift-json need use version for scala 2.11.

/home/rockiey/git/spark-highcharts/target/spark-highcharts-0.6.4.jar

net.liftweb:lift-json_2.11:2.6.3

Select Data to Plot

Chart Options

Demos

Type Mapping between Scala and Highcharts

Scaladocs

download spark-highcharts and lift-json jar file.

add following line to spark-defaults.conf

spark.jars spark-highcharts-0.6.4.jar,lift-json_2.10-2.6.3.jar

spark-highchart use Apache 2.0 License

However, the Highcharts JavaScript library that is included in this package is not free for commercial use. Please contact Highcharts for license related issues.