Use Bokeh with IPython Notebook
You can use Bokeh to create interactive graphs. You will be able to graphically represent your data in Python alone, including output in HTML and integration with IPython Notebook.
For how to use it, refer to the gallery on the official website and the notebook below.
- Gallery — Bokeh 0.7.0 documentation
- Bokeh 5-minute Overview
- Creating Interactive Visualizations with Bokeh
The main developers of Bokeh are those from Continuum Analytics who provide Anaconda and others, and you can also see the author's presentation on YouTube. It's almost 50 minutes long, but it's helpful because there are also architectural topics such as BokehJS.
As for the version, the latest version was released last week. The content is described in the release article-Bokeh 0.7 Released!-. It's not stable yet, such as adding new features or changing APIs, but I think it's very useful if you want to explore the data. Recently, implementations of Scala and Julia have been added, and it seems that they can be used in processing systems other than Python.
For Python, you can easily install it with pip. However, I think that it is often used in combination with NumPy and Pandas, so I added it to the Docker image. With skitazaki / python34-ipython you can use some scientific computing libraries with Bokeh in your IPython Notebook.
Since functions such as zooming are effective in the graph part, it is an advantage over static graphs that you can roughly plot and then adjust. It is also possible to download it as an image when the display is just right with zooming.
It will be a useful tool for visually viewing affordable datasets.
The image of Notebook usage example is as follows. Since the source code, execution results, and explanations can be stored together, it seems good to introduce it by saving notes.
Recommended Posts