A collection of ipython notebooks I've made for various projects, with nbviewer links:
Scripts to make maps of Canadian Centre for Climate Modeling and Analysis climate models to the year 2100: CCCma_Basemap.ipynb Script to visualize one such map CCCma_SNC_to_2099.ipynb Script to extract a feature on the same day per year and make an animated GIF
earthquakes-jp.ipynb and earthquakes-rest-api.py: Scripts to build an animated GIF of earthquakes in China and Japan. See the final result at http://www.prooffreader.com/2014/06/Japan_Earthquakes/animated-map-of-earthquakes-near-china.html
Scripts to analyze Billboard magazine pop charts data: billboard_top_words.ipynb : script to find the most decade-specific words in song titles since 1890.
Scripts to analyze word use in Star Trek: The Original series: munge_tos_transcripts.ipynb: script to download fan scripts, create a transcript database and a bokeh plot of dialogue per character
Scripts to analyze CDC's mortality files: unusual_mortality.ipynb : script to show ten unusual causes of death in the CDC mortality database.
monty_monte.ipynb: A Monte Carlo simulator of the Monty Hall problem! You can choose the number of doors and of goats, and the number of iterations.
pocket_tumblr_reddit_api.ipynb: A script to search the hard drive, Pocket or Tumblr for photos, then save them or upload them to Tumblr or Reddit. The Reddit posting writes to a json on an sftp server, where a cron job can run the Reddit/PRAW script once an hour.
top_10_python_idioms.ipynb: A notebook showing the top 10 idioms I wished I'd internalized earlier when I was first learning Python in 2012 coming from mostly Visual Basic (ugh, I know!).
Scripts to convert tree/graph files between a few formats for use in MEGA or Gephi: (1) tree_convert_mega_to_gexf.ipynb; (2) tree_convert_mega_to_json.ipynb; (3) tree_convert_newick_to_json.py
weather_ML.ipynb: a script to analyze a huge (7,000,000-row) table of weather data that had lost its column identifiers and changed column order frequently. I used conservative criteria to assign unambiguous columns (e.g., only rainfall had numbers in the 0.00-0.10 range, snowfall was NULL during July), and then used machine learning (random forest classifier) to classify the rest based on a handful of metrics, like whether values were correlated with time of year or time of day, how many had a value of zero, whether they were integers or floats, their medians and standard deviations, etc. The source data is proprietary, so is not included here.