MCMC Results Saving System Feature Requests

Feature requests

• Log the MCMC parameters (and values for TDCFs) stored in a log file / HDF5 file. If re-run multiple times, the log file would have timestamps for data values recorded and append the latest run to the file
• Fudge factor for certain things <-- need clarification on this from Joe
• Data file format and a function that reads the data into a python class at run time, with appropriate query functions, not code the python structure into the data file which makes it extremely difficult to change or scale things.
• Consider backwards compatibility with already existing actProd, sensProd type scripts
• Imagining the future work flow, managing time stamps of MCMC runs may get intractable given that we often run MCMC on a single measurement multiple times, and/or if we find a flaw (likely months later) in the model elements that are divided out from the IFO measurement prior to input to the MCMC, or we're running "the final analysis" to get all our MCMC posterior ducks in a row for systematic error budgeting, we're gunna have a bad time keeping track of which time stamp is the right time stamp. Consider the option to overwrite a given entry, but it's OK to say "the .json" file is just a text file. It's up to the user to manage it at that level."