mwaskom/seaborn ? reverse-engineered prompt

Build me a Python library for statistical data visualization, something like a clean high level wrapper around matplotlib that makes common plots easy and attractive.

I want it to work nicely with pandas data frames, support the usual kinds of charts people use for data analysis, and include a few more advanced statistical plots when the extra scientific packages are available. Keep the public API simple and friendly, with solid defaults so users can make good looking plots without a lot of code.

Please include a small set of examples and a proper test suite, and make sure the code follows a consistent style. If you need to check current best practices for any plotting or packaging details, look up the latest docs online. The end result should feel polished enough for real data science work, and be easy to install and use in Python 3.10 or newer.