1. Deprecate the convert feature in place of a more general "save"
2. Fix the result parse in method to accept either positional or optional
3. Force the user to pick a single option using a mutually exclusive parser
4. Add an option to fix the maximum number of samples
5. Add a lightweird option to strip down the results
6. Improve the evidence printing output
7. Minor adjustments to core internals to help debugging

Fix error in Categorical prior and add tests

See merge request lscsoft/bilby!990

Fix MCMC tests

See merge request !985

Add a categorical prior

See merge request !982

Remove the autoburnin causing kombine to fail the CI tests

Closes #585

See merge request !988

bilby/gw/likelihood.py: speed up the spline interpolation in ROQ

See merge request !971

Replace bessel interpolant to scipy function

See merge request !976

Improve checkpoint stats plot

See merge request !977

Add cdf method to PriorDict classes

See merge request !943

Fix typo in sampler documentation

See merge request !986

Add a built-in mcmc sampler

See merge request !905

Fixes issue that causes ConditionalDeltaFunction posterior samples not to be saved correctly

See merge request !973

Solution to Incorrect Injected SNR values in log files

See merge request !980

Make Python 3.6+ a specific requirement

See merge request !979

Fix to the calibration and time marginalized likelihood

See merge request !978

Resolve "Distance marginalization error"

Closes #571

See merge request !960

Resolve "check_draw doesn't catch nan log_likelihood values."

Closes #559

See merge request !965

Resolve "docs configuration overwrites source suffixes"

Closes #574

See merge request !963

Resolve "Lack of documentation on class side-effects"

Closes #572

See merge request !964

Allow ptemcee initialization with array

See merge request !955

Resolve "Get rid off `Prior.test_valid_for_rescaling`?"

Closes #511

See merge request !956

Add run statistics to the dynesty meta data

See merge request !969