serialization - PyMC change backend after sampling -


I am using PyMC in the analysis of some high energy physics data. It has worked for perfection, the analysis is complete, and we are working on the paper.

I have a small problem, though. I ran the sample with the RAM database backend in the IAIthon kernel process, for a few months this mark is sitting around in memory. The problem is that the workstation support staff want to upgrade a kernel and want to reboot the workstation. This will lose the mark. I want to keep these marks (as opposed to just creating new ones) because they are those which I have made all the plots. I would like to include a part of the mark (only the parameters of interest) as the supplementary material with publication.

Is it possible to take an existing series in pymc.MCMC Object created with backend, change in a different backend, and mark the series?

Trace values ​​are stored in the form of nappace arrays, so you have the value of each parameter in the file You can use numpy.savetxt to send. (This text is the backend that comes under the hood.)

It is a good idea to save your current mark, I suggest that before publishing your analysis takes the time to repeat.


Comments

Popular posts from this blog

ios - How do I use CFArrayRef in Swift? -

eclipse plugin - Run java code error: Workspace is closed -

c - Error on building source code in VC 6 -