I'm trying to run GraphX on Ipython notebook. Firstly, I launched Spark/Hadoop clusters and then launched ipython notebook using this tutorial (http://blog.insightdatalabs.com/jupyter-on-apache-spark-step-by-step/). But now I have only Python 2 kernel and I need GraphX to work with graph. How can I run it?
0
Add a comment
|
1 Answer
GraphX can only be referenced by the Scala API. Saying this you can either use something like Jupyter-Scala so you can use Scala within Jupyter or you can use GraphFrames which supports the Python API. For more information on GraphFrames, please check out Introducing GraphFrames.
6 Comments
Alex Ermolaev
I wanted to install apache toree instead but got the following error: No such file or directory: '/usr/local/spark/python/lib'
Denny Lee
This may be a case where the default location where Apache Toree is expecting the PYTHON_PATH isn't where you had installed python. Could you check the toree kernel.json and update the "PYTHONPATH": "/usr/local/spark/python:..." to your location. For more information, please check technippet.blogspot.com/2016/02/…
Alex Ermolaev
I followed this tutorial: launched apache toree kernel, but it doesn't respond (
Denny Lee
Could you provide some more context as there isn't enough information here to debug. Saying this - why not use GraphFrames so that way you can use Python?
Alex Ermolaev
yeah, I think it would be great to use GraphFrames, but how can I integrate it with Ipython? Now in my /usr/local/share/jupyter/kernels I have only apache_toree_scala.
|