Ibis on Impala: Python at Scale for Data Science

Ibis on Impala: Python at Scale for Data Science by Marcel Kornacker and Wes McKinney.

From the post:

Ibis: Same Great Python Ecosystem at Hadoop Scale

Co-founded by the respective architects of the Python pandas toolkit and Impala and now incubating in Cloudera Labs, Ibis is a new data analysis framework with the goal of enabling advanced data analysis on a 100% Python stack with full-fidelity data. With Ibis, for the first time, developers and data scientists will be able to utilize the last 15 years of advances in high-performance Python tools and infrastructure in a Hadoop-scale environment—without compromising user experience for performance. It’s exactly the same Python you know and love, only at scale!

In this initial (unsupported) Cloudera Labs release, Ibis offers comprehensive support for the analytical capabilities presently provided by Impala, enabling Python users to run Big Data workloads in a manner similar to that of “small data” tools like pandas. Next, we’ll extend Impala and Ibis in several ways to make the Python ecosystem a seamless part of the stack:

  • First, Ibis will enable more natural data modeling by leveraging Impala’s upcoming support for nested types (expected by end of 2015).
  • Second, we’ll add support for Python user-defined logic so that Ibis will integrate with the existing Python data ecosystem—enabling custom Python functions at scale.
  • Finally, we’ll accelerate performance further through low-level integrations between Ibis and Impala with a new Python-friendly, in-memory columnar format and Python-to-LLVM code generation. These updates will accelerate Python to run at native hardware speed.

See: Getting Started with Ibis and How to Contribute (same authors, opposite order) in order to cut to the chase and get started.


Comments are closed.