Seaborn is a library of high-level functions that facilitate making informative and attractive plots of statistical data using matplotlib. It also provides concise control over the aesthetics of the plots, improving on matplotlib's default look.

Online documentation is available here.

There are a few tutorial notebooks that offer some thoughts on visualizing statistical data in a general sense and show how to do it using the tools that are provided in seaborn. They also serve as the primary test suite for the package. The notebooks are meant to be fairly, but not completely comprehensive; hopefully the docstrings for the specific functions will answer any additional questions.

Controlling figure aesthetics in seaborn

Plotting complex linear models

Visualizing distributions of data

Plotting statistical timeseries data

• Python 2.7

• numpy

• scipy

• matplotlib

• pandas

• statsmodels

• patsy

• scikit-learn

• husl

• moss

To install the released version, just do

pip install seaborn

However, I update the code pretty frequently, so you may want to clone the github repository and install with

python setup.py install

from within the source directory.

To test seaborn, run make test in the source directory. This will execute the example notebooks and compare the outputs of each cell to the data in the stored versions. There is also a (small) set of unit tests for the utility functions that can be tested separately with nosetests.

https://github.com/mwaskom/seaborn

Please submit any bugs you encounter to the Github issue tracker.

"Those are nice plots" -Hadley Wickham