Support Questions

Find answers, ask questions, and share your expertise

Pyspark Interpreter not working on Zeppelin

New Contributor


I'm using HDP sandbox 2.6.4, with Zeppelin Notebook installed.

When I want to use Pyspark on Zeppelin, it won't work...

Example :

%pyspark print "Test"


Traceback (most recent call last): File "/tmp/", line 302, in <module> __zeppelin__._setup_matplotlib() File "/tmp/", line 141, in _setup_matplotlib import backend_zinline File "/usr/hdp/current/zeppelin-server/interpreter/lib/python/", line 30, in <module> import mpl_config File "/usr/hdp/current/zeppelin-server/interpreter/lib/python/", line 99, in <module> _init_config() File "/usr/hdp/current/zeppelin-server/interpreter/lib/python/", line 83, in _init_config fmt = matplotlib.rcParams['savefig.format'] KeyError: 'savefig.format'

I can't cancel the execution...

And on the Resource Manager UI, the job is running indefinitely : (See attached png file)

Thank you for your help



New Contributor

According to this JIRA :

The issue is the version of the package matplotlib. I've got this version : but the minimum version required is 1.2.x

With pip I can't upgrade the version because of the version of python, that is 2.6.6 and so depreciated.

New Contributor

Problem solved !

First : Install Python 2.7 using this tuto :

Second : Install matplotlib with python2.7 : python2.7 pip install matplotlib

Third : Configuring the new version of Python as default for Spark in Zeppelin using this tuto :

Now It works !


Expert Contributor

And after Python 2.7 installation don't forget to change Zeppelin Spark interpreter setting as: