在 emr 集群上安装 com.databricks.spark.xml

Question

Does anyone knows how do I do to install the com.databricks.spark.xml package on EMR cluster.

I succeeded to connect to master emr but don't know how to install packages on the emr cluster.

code

sc.install_pypi_package("com.databricks.spark.xml")

Answer 1

On EMR Master node:

cd /usr/lib/spark/jars
sudo wget https://repo1.maven.org/maven2/com/databricks/spark-xml_2.11/0.9.0/spark-xml_2.11-0.9.0.jar

Make sure to select the correct jar according to your Spark version and the guidelines provided in https://github.com/databricks/spark-xml .

Then, launch your Jupyter notebook and you should be able to run the following:

df = spark.read.format('com.databricks.spark.xml').options(rootTag='objects').options(rowTag='object').load("s3://bucket-name/sample.xml")