<img src="images/continuum_analytics_logo.png" alt="Continuum Logo", align="right", width="30%">

Building Python Data Applications
with Blaze, Numba, and Bokeh

NCCU PyDay 2015

by Andy Terrel

Tutorial sections¶

• Section 0: Introduction
• Section 0a: Numba First Steps
• Section 0b: Numba Types
• Section 0c: Numba and Numpy
• Section 1: Blaze
• Section 2: Bokeh
• Section 3: Apps

Follow along¶

http://git.io/vIlk2¶

• Option A: Download repository

https://github.com/aterrel/pydata-apps¶

• Option B: View notebooks on nbviewer

http://nbviewer.ipython.org/github/aterrel/pydata-apps¶

Numba¶

http://numba.pydata.org

Numba gives a simple way to compile simple functions to assembly via llvm to be used with NumPy kernels.

Blaze¶

http://blaze.pydata.org/en/latest/¶

Blaze allows Python users a familiar interface to query data living in other data storage systems.

Bokeh¶

http://bokeh.pydata.org/en/latest/¶

Bokeh is a Python interactive visualization library that targets modern web browsers for presentation.

Tutorial goals¶

• Learn basics of Numba
• Query backends with Blaze expressions
• Generate Bokeh visualizations
• Create data applications with Blaze and Bokeh

Installation¶

• Download and install the Anaconda Python Distribution

• Download archive of this repository or checkout with git

  git clone https://github.com/aterrel/pydata-apps.git
  

• Each tutorial has a slightly different set of requirements. To download all the requirements try:

  conda update conda
  conda env create
  

• Activate the environment

  source activate pydata_apps
  

Static notebooks¶

For those want to just follow a static notebook (not all interactive elements will work), see the following links:

• Section 0: Introduction
• Numba 0: Numba First Steps
• Numba 1: Numba Types
• Numba 2: Numba and Numpy
• Section 1: Blaze Tutorial
• Section 2: Bokeh Tutorial