COSMIC-PopSynth/COSMIC: COSMIC paper Code Release (Draft Part 2)

This is the version that is used to produce results in the COSMIC paper: https://arxiv.org/abs/1911.00903

We have adjusted the way we handle mass loss due to a SN.

We have adjusted the way that the output time-resolution is handled. A user can now ask for different time resolutions during difference stages of the evolution. See the docs for examples.

We have added the ability to restart a binary in the middle of the evolution. A user can select a point in the bpp or bcm array and re-run the evolution to the end, getting identical results (with the caveat of restarting after one of the objects has undergone a SN then the evolution is not perfectly re-produced. See docs for an example of re-starting the evolution

We have updated the sampling code to include (verified) sampling methods for p_orb, ecc, masses for the independent sampler method (the multidim method is unchanged), There are plots in the docs highlighting the resulting distributions.

We have changed the way we report kick information for the evolution. It is now its own returned output from running Evolve.evolve. So you now get the initC, bpp, bcm, kick_info. This saves a lot on memory as the kick_info array is a 2 dimensional array with the following properties, see the docs as well.

• kick_info[i,1]: snstar of exploding star
• kick_info[i,2]: disrupted (0=no, 1=yes)
• kick_info[i,3]: magnitude of the natal kick
• kick_info[i,4-5]: phi and theta (in the frame of the exploding star)
• kick_info[i,6]: eccentric anamoly
• kick_info[i,7-9]: change in 3D systemic velocity of the binary, or the change in 3D velocity of snstar=1 if the system is disrupted
• kick_info[i,10]: magnitude of systemic velocity of the binary if bound or magnitude of total velocity of snstar=1 if disrupted, accounting for both SNe
• kick_info[i,11-13]: change in 3D velocity of the snstar=2 if system is disrupted
• kick_info[i,14]: magnitude of velocity of snstar=2 if disrupted, accounting for both SNe
• kick_info[i,15]: (total) tilt of the orbital plane after each SN w.r.t. the original angular momentum axis after each SN
• kick_info[i,16]: azimuthal angle of the orbital plane w.r.t. spins
• kick_info[i,17]: random seed at the start of call to kick.f

Finally, we now update the natal_kick_array that is reported in the returned initC table from running Evolve.evolve with the following parameters updated to what they actually were during the evolution of the binary (this also has two rows one for each SN)

NATAL_KICK_COLUMNS = ['natal_kick',
                      'phi',
                      'theta',
                      'eccentric_anomaly',
                      'randomseed']
`