kima-run and several new features

This is the fourth official release, a bit later than usual.

One major new addition is the kima-run command now included in pykima to compile and run kima jobs.
This replaces the old make; ./run combo, and offers some more functionality.

Other new features:

• kima-checkpriors now accepts a column name in addition to a column number
• new interface to build models within Python using pykima.KimaModel
• implementation of the "known object" mode for Keplerians with individual priors for the orbital parameters
• the plot with the posterior for orbital periods now also shows the prior
• dedicated test suite using some of the examples
• an even bigger list of examples demonstrating most of kima's features
• header of sample.txt now matches the number of columns
• new option for up to cubic trends (#60)
• re-implemented the Student-t likelihood (#61) with automatic outlier "detection"
• allow for a quasi-periodic celerite kernel, created by Andrew Collier Cameron
• the obs_after_HARPS_fibers has been discontinued as it was error-prone is replaced with the multi-instrument mode (#65)
• the kima-gui script is included but still needs further testing

kima v3

This is the third official release.
Analysis fo data from different instruments is substantially improved. Added possibility of linear correlations with activity indicators and moving average models. New syntax to define priors and much simpler kima_setup.cpp files.
The documentation in the wiki is updated to reflect these changes. The changelog will be updated shortly.

kima v2

This is the second official release.
The main new feature is the ability to analyse RVs from different instruments.
kima will automatically add new parameters for the RV offsets between instruments and extra jitters for each instrument. The documentation in the wiki is being updated to reflect this change. Also see the changelog and the updated roadmap.

First release of kima

This is the first official release.
kima has been accepted for publication in JOSS (see journal page, review).