Blogging with Jupyter

I realized that most of my blog posts behave like a jupyter notebook (markdown, code and repeat). So why not just write a jupyter notebbok and post it as a blog entry. I am going evaluate if this is as nice as writing with prose.io.

How it works

1. git clone the repo containing the jekyll blog
2. jupyter notebook run the notebook and write stuff
3. ipython nbconvert <your_notebook>.ipynb --to markdown
4. Move the generated markdown to the “_posts” folder Caution: The images of this notebook are placed into a folder called <your_notebook>_files.

%matplotlib inline

import numpy as np
import matplotlib.pyplot as plt


plt.rcParams['figure.figsize'] = (14, 6)

x = np.linspace(0.0, 20.0, num=100)
y1 = np.sin(x)
y2 = np.sin(x*2)
y3 = 1.5 * np.sin(x)
y4 = 0.8 * np.sin(0.8*x + 1)

plt.plot(x, y1, 'k--', x, y2, 'g--', x, y3, 'b--', x, y4, 'r--')

[<matplotlib.lines.Line2D at 0x7fed42f93710>,
 <matplotlib.lines.Line2D at 0x7fed42f937d0>,
 <matplotlib.lines.Line2D at 0x7fed42f93fd0>,
 <matplotlib.lines.Line2D at 0x7fed42f9d410>]

Tricks with jupyter notebook

• get a plot.ly account and generate interactive graphs
• use %load_ext autoreload %autoreload 2
• display multiple beatiful outputs (not only the last of a cell): from IPython.core.interactiveshell import InteractiveShell InteractiveShell.ast_node_interactivity = "all"
• access the docs by prepending a question mark and evaluating a cell: ?str.replace
• list all ipython magic commands: %lsmagic
• use variables in multiple notebooks like global context: %store data and %store -r data (list all with %who)
• %%time time information of a cell about a single run
• %%timeit runs the same cell 10.000 times and returns the mean of the durance
• run shell commands: !ls -la or !python3 -m install numpy
• LaTeX formulas: \\( P(A \mid B) = \frac{P(B \mid A) \, P(A)}{P(B)} \\) \(P(A \mid B) = \frac{P(B \mid A) \, P(A)}{P(B)}\)

(More ideas: jupyter notebook tips)