A little word cloud generator in Python.

Needs PIL, numpy and Cython (>= 0.16).

The example uses scikit-learn for extracting word counts from a text. For scikit-learn <= 0.11, you have to remove the min_df keyword.

Build using make or python setup.py build_ext -i.

If you are not on Ubuntu, you need to adjust FONT_PATH to point to some existing font.

See my blog for some details: http://peekaboo-vision.blogspot.de/2012/11/a-wordcloud-in-python.html

For the blog post I removed the word shall from the constitution word count which is not in the scikit-learn stopword list.