timing model sampling

[siyuan-chen]

Hi all,

there is a bug in timing.py , where if your libstempo pulsar object has unsorted TOAs/residuals, you will subtract very different numbers from the enterprise pulsar object. Not sure, if there are any other bugs with the tming model parameter sampling, but I am testing it on a few different pulsars from the EPTA and comparing against the temonest equivalent posteriors.

For now, the small sorting issue bug has been fixed and can be merged through pull request #59

Cheers