Pyproffit

- Easy-to-use MCMC implementation for any built-in or user-defined model, see this [tutorial](https://pyproffit.readthedocs.io/en/latest/Discontinuities.html#Running-Monte-Carlo-Markov-Chain)
- Prior probabilities can be set either automatically or through any custom user-defined function
- Corner plots and posterior distributions
- Posterior envelopes can be automatically displayed together with the data
- A few bug fixes

Updates in iminuit (>2.0.0) introduced a breaking interface between pyproffit and iminuit. This version is now compatible with iminuit >2.0.0.

New features:
- Azimuthal scatter computation as a new method in the "Profile" object
- The median surface brightness profile can now be extracted in elliptical annuli; it also returns bootstrap samples and covariance matrix to take the cross-talk between bins into account
- The fit now displays best-fit chi-square/C-stat for evaluation of goodness of fit
- Output model images can now be extracted directly from the loaded surface brightness profile to get a model of the azimuthally averaged emission

This is the version of the code released together with the Eckert et al. (2020) paper