Hide the input cells from your IPython slides
A tweet arrived...
@damian_avila Is there a way to only show the output of a cell in an NB slideshow?
— Thomas Wiecki (@TWiecki) October 10, 2013
And my answer was YES...
How we can easily do it? Just follow these few steps:
A simple minimap for your nbconverted ipynb
Some days ago, I presented a way to mimic the execution of nbconverted IPython notebooks (ipynb). Today, I will show you how to add a simple minimap to these static html documents.
But, what is a minimap? Do you know SublimeText or Ninja-IDE? They are IDEs with a thin column to the right (of the main window) which provides an overview of the code we are working on... OK, you don't have an idea what am I talking about? Just go ahead and you will see...
Mimic the IPython notebook cell execution
Yesterday, I was visiting the Jake VanderPlas's Scientific Python Course site because of some issues raised at the ipython-dev mailing list (and obviously, because Jake usually do amazing things).
To build his site, Jake wrote an Hyde plugin to render ipynb files to static html files using the IPython.convert platform. You can see an example of the final rendered document here.
When I saw the final document, I wondered if we could have hidden output cells which automatically get visible after a click over the input cell... and then, I made some experiments using the power of the IPython.nbconvert library ;-)
Nikola's Zen theme finally released
Just a quick update... Some weeks ago I blogged about my new Nikola's theme family called Zen. In fact, you are seeing the zen-ipython theme powering my blog right now.
I have used them for several days without any remarkable issue, so I decided to release them at the Nikola's theme repository.
There, you will see three Zen themes:
-
zen, based in mako templates -
zen-jinja, the same zen theme but based (obviously) in jinja templates -
zen-ipython, a jinja-based zen theme suited to be used with the IPython notebooks.
vIPer is now using IPython.nbconvert
As you know (if not, probably the following lines gives you an idea), vIPer was originally designed to export from ipynb files to:
- a plain static html
- a slideshow powered by Landslide
With the last release, IPython support this kind of exportation using the well designed and easily extensible IPython.nbconvert library. Now, it is easier to incorporate the IPython machinery inside vIPer and support the exportation to several formats.
Change the IPython slides defaults with an IPython config file
Some people are asking me how to change the IPython slides's defaults. You can see the description of the request here. Essentially, he wants to change the default transition between the slides.
To solve this question, we introduce the notion of IPython config files, which let us easily change things inside IPython.nbconvert (the library where the IPython slides lives) to achieve our desires, at least in the slideshow generation issue ;-)
First of all, you need to know some details:
Using a local Reveal.js library with your IPython slides
IPython slides are powered by Reveal.js library, you probably already know that.
But... you probably don't know that we use jsdelivr CDN to load all the necessary js and css files. As a consequence, you don't need to download Reveal.js, but also, the speaker notes are not working by default.
But... what happens, if you have low connectivity in the place where you will be presenting your talk (or if you want to use the speaker notes feature)?
Different ways to lead an open source project
Just a few lines, to trigger the discussion... to describe my point, I will pick two projects I have contributed (and I want to contribute even more): IPython and Nikola.
Some months ago, Brian Granger, a core developer from the IPython project gave a talk at SciPy 2013 where he said:
A new Nikola's theme for my site: Zen
Customizing your IPython slides
Some days ago, I received this message on G+:
In a future release, would be handy to add custom classes to slides so that you can use CSS rules to choose certain layouts for image slides, text, multiple charts, etc.
But... we already support this kind of customization! And it is clearly my fault that you don't know about this feature... so: